1. food and drink
2. healthy eating

ADOLESCENTS’ HEALTHY EATING
– A SOCIO-COGNITIVE APPROACH TO
STUDYING CONSUMER SOCIALISATION
AND BEHAVIOUR CHANGE
By Susanne Pedersen
PhD thesis submitted to
School of Business and Social Sciences, Aarhus University,
in partial fulfilment of the requirements of
the PhD degree in
Business Administration
February 2015
ACKNOWLEDGEMENTS
My time as a PhD student has come to an end, and I can look back at some
exciting years filled with many challenges and good experiences – and a lot of hard work.
I want to acknowledge all those good people surrounding me, since this PhD thesis would
never have been possible without the assistance and valuable support from several.
First and foremost, I am extremely grateful to my supervisors John
Thøgersen and Alice Grønhøj for their guidance, support, continued encouragement and
positive approach to me and my work. John, thanks for being a demanding and kind
supervisor. You always had time for questions, discussions, feedback and for sharing
your massive knowledge on statistics and social psychology. Alice, thanks for being a
caring, detailed-oriented supervisor sharing my research interest in children and
adolescents. You encouraged me to pursue a PhD and made me feel comfortable with
doing research.
This doctoral research would not have been possible without the funding of
the research project “Step by step changes of children’s preferences towards healthier
food”, by the Danish Ministry of Science, Technology and Innovation and the MAPP
Centre at Aarhus University as co-financer. I would like to express my gratitude to
project manager Tino Bech-Larsen and project group members Alice Grønhøj and Jessica
Aschemann-Witzel for your support, good cooperation and insightful discussions. Also
thanks to research assistants and the wider project group for inspiration and cooperation.
I want to thank colleagues in all functions at MAPP and BADM for sharing
your expertise and for helping and supporting me throughout the years. Special thanks to
Birgitte Steffensen and Karin Hørup for proofreading my dissertation.
i
During my second year as a PhD student, I had the opportunity to stay at
Department of Psychology at Bath University for three months. I am very grateful to
Professor Bas Verplanken for showing me great hospitality and for introducing me to the
department and to one of his research groups. I also want to thank Deborah Roy and
Aarhen Knight Ahumada for intense discussions and enjoyable data collections. I truly
enjoyed the stay!
I also want to thank my assessment committee: Professor Bas Verplanken,
Professor Lotte Holm and Professor Liisa Lähteenmäki for valuable comments and
recommendations helping me to improve my work.
My PhD study coincided with some very unfortunate and critical events in
my life. It has delayed my thesis, but I still hand in on time. I want to thank AU and the
PhD school at BSS for making it possible to delay this thesis. Also warm thanks to all at
BADM who have cared and supported me in what were very difficult times for me and
my family.
I also want to express my warmest gratitude to Berit Kamp Kragh who has
been so much more than a colleague sharing an office. You have been a true friend
supporting me in every possible way sharing both tears and laughter - for that I thank you.
To my parents, family and friends in Western Jutland, my brother and sisterin-law in China, my Icelandic family in Iceland and Norway – thanks for all your support
throughout the years. I also want to thank my friends – especially Edith, Bente, Lisette,
Birgitte, Ida and Lise for your continuous support in all aspects of my life.
To my dearest trio, Selma, Rakel and August, I want to thank you for your
straightforwardness in questioning my work and for demonstrating the challenges of
ii
family interaction and consumer socialisation regarding healthy eating. I hope we are
doing things right! Most importantly, I thank you for always reminding me about the
important things in life.
Last, but certainly not least, I want to express my deepest gratitude to my
dear husband Óskar. You are my bridge over troubled waters, my rock, my Viking. You
stand by my side in good times as well as in bad. Through your love, patience, support
and belief in me, I find the energy to carry on and do my best. Thank you.
Aarhus, May 11, 2015
iii
“The unique nature and importance of adolescence mandates explicit and
specific attention in health policy and programmes”
(WHO, 2014)
iv
RESUMÉ (SUMMARY IN DANISH)
Denne ph.d. afhandling bidrager med viden om unges adfærd i forhold til at
spise sundt. I afhandlingen undersøges forbrugersocialisering samt sociale normers
påvirkning og muligheder for adfærdsændringer i forhold til unges sunde kostvaner. I
introduktionen begrundes vigtigheden af at studere unge og sunde kostvaner, og der
redegøres for, at en mere holistisk tilgang er nødvendig for at imødegå den stigende
overvægt blandt unge. Det er vigtigt at forstå udviklingen i og påvirkningen af unges
kostvaner og muligheden for at promovere sunde kostvaner gennem interventioner. Ved
at gennemgå relevant litteratur om forbrugersocialisering, social påvirkning og
adfærdsændring ved hjælp af feedback-interventioner i forhold til unges kostvaner
argumenteres der for, at en social-kognitiv tilgang til forbrugersocialisering og
adfærdsændring giver en dybere og mere nuanceret forståelse af unges kostvaner. Baseret
på dette præsenterer denne afhandling tre forskningsspørgsmål, som behandles i tre
artikler. Forskningsspørgsmålene er:
1. Hvilke roller har forældre og unge i forbrugersocialiseringen af de unges sunde
kostvaner?
2. Hvilken betydning for unges sunde kostvaner har den sociale påvirkning fra forældre
og venner sammenlignet med personlige faktorer?
3. Hvordan kan en intervention baseret på social-kognitiv teori og sms-baseret feedback
forbedre unges sunde kostvaner og hvorfor?
Det første forskningsspørgsmål besvares i den første videnskabelige artikel.
Fordi familiens interaktion og familiemedlemmers rolle i socialiseringen af sunde
kostvaner er underbelyst, havde dette studie til formål at undersøge unge og deres
forældres bevidsthed om og involvering i sunde kostvaner i relation til deres roller i
v
forbrugersocialiseringen af sunde kostvaner i familien. Som en opfølgning på en
intervention om sunde kostvaner deltog 38 unge og deres familier i dybdeinterviews og
en praktisk øvelse om dagligt anbefalet frugt- og grøntindtag. Resultaterne viste, at de
unge benyttede sig af to strategier: en direkte strategi med krav til forældrene eller en
samarbejdsstrategi, som hjalp forældrene. Forældrene påbegyndte samtaler med den unge
og andre familiemedlemmer om sunde kostvaner og følte sig ansvarlige som
rollemodeller og efterkom ofte den unges krav og anerkendte den unges hjælp.
Resultaterne bekræfter, at forældre stadig har afgørende betydning i forhold til den unges
sunde kostvaner, men med den unge som en aktiv spiller. Det bekræfter, at
forbrugersocialisering kan opfattes som bi-direktionale processer. Studiet supplerer
tidligere studier ved at inkludere den unges nærmeste familie som analyseenhed. Med en
helhedsorienteret tilgang til at studere forbrugersocialisering bidrager denne artikel til at
identificere og forstå barrierer og katalysatorer for unges sunde kostvaner.
Det andet forskningsspørgsmål besvares i den anden videnskabelige artikel.
Formålet med artiklen er at teste, om den almindelige opfattelse, at børn i stigende
omfang bliver påvirket af venner på bekostning af forældre i puberteten, også gælder for
sunde kostvaner. Spørgeskemaer blev udfyldt af 757 dyader bestående af unge og deres
forældre. Artiklen tager udgangspunkt i social kognitiv teori og The Focus Theory of
Normative Conduct og finder, at forældre forbliver dem, som primært påvirker unges
sunde kostvaner, hvor det de gør (deskriptive normer), er vigtigere, end hvad de siger
(injunktive normer). Studiet bidrager til en bredere forståelse af, hvad der påvirker unges
sunde kostvaner ved at sammenligne påvirkningen fra flere sociale sfærer (forældre og
venner) samtidig med, at der kontrolleres for personlige faktorer såsom unges tiltro til
egne evner (self-efficacy) og forventninger til resultatet (outcome expectations). Tidligere
studier har ikke inkluderet alle disse faktorer i den samme analyse. Implikationerne af
vi
studiet er, at (1) interventionen rettet mod sunde kostvaner bør have til formål at styrke
tiltroen til egne evner og styrke de positive forventninger til resultatet blandt unge, (2)
familiekonteksten bør inkluderes, når interventioner om sunde kostvaner implementeres,
og (3) forældres bevidsthed om deres påvirkning af unges sunde kostvaner bør styrkes.
Det tredje forskningsspørgsmål besvares i den tredje videnskabelige artikel.
Artiklen undersøger effekterne af en intervention rettet mod sunde kostvaner, der ved
brug af feedback via sms over 11 uger havde til formål at forbedre unges adfærd, tiltro til
egne evner og positive forventninger til resultatet i forhold til at spise frugt og grønt.
Spørgeskemaer før og efter interventionen blev udfyldt af 1488 unge, som vilkårligt blev
allokeret til en kontrolgruppe og to eksperimentgrupper. Begge eksperimentgrupper satte
ugentlige mål for indtag af frugt og grønt, rapporterede deres indtag dagligt og dernæst
modtog feedback på deres indsats – alt sammen via mobiltelefonens sms-funktion. Den
ene eksperimentgruppe modtog derudover 45 minutters undervisning om ernæring af en
diætist i skolen. De direkte effekter af interventionen var ikke signifikante, men for unge,
der deltog i sms-rapporteringen, var der signifikante effekter for niveauet af deltagelse i
interventionen afspejlet i antallet af sendte sms’er. Deltagere, der sendte flere end
halvdelen af de mulige sms’er, øgede deres indtag af frugt og grønt signifikant. Deltagere,
der sendte mellem 10 og 50 % af de mulige sms’er, oplevede et signifikant fald i tiltroen
til egne evner i forhold til at spise frugt og grønt, og de, der sendte færre end 10 % af de
mulige sms’er, oplevede et signifikant fald i positive forventninger til resultatet af at spise
frugt og grønt. Resultaterne indikerer, at deltagernes aktive engagement i en intervention
er afgørende for dens succes. Dette medfører implikationer for design og udførelse af
sundhedsfremmende interventioner.
vii
viii
EXECUTIVE SUMMARY
This PhD thesis contributes with knowledge about adolescent healthy eating
by studying consumer socialisation, social influence and behavioural change in relation to
adolescent healthy eating. The introduction provides the important reasons for studying
adolescents and healthy eating and explains that a more holistic approach is needed in
order to respond to the rising levels of overweight among adolescents. It is important to
understand the development of and influences on adolescent healthy eating behaviour and
the possibilities for promoting healthy eating through interventions. By reviewing
relevant literature on consumer socialisation, social influence and behaviour change
through interventions employing feedback in relation to adolescent healthy eating, it is
argued that a socio-cognitive approach to consumer socialisation and behaviour change
provides a richer and more nuanced understanding of adolescent healthy eating. Based on
this, the thesis presents three research questions which are investigated in three research
papers. The research questions are:
1. Which roles do parents and adolescents have in healthy eating socialisation?
2. How does the social influence from parents and friends compared to personal factors
impact adolescents’ healthy eating?
3. How can a feedback intervention based on socio-cognitive theory and using text
messaging improve adolescent healthy eating and why?
The first research question is answered in research paper 1. Since the area of
family interaction and family members’ roles regarding healthy eating socialisation is
underexposed, the study aimed at exploring adolescents’ and parents’ awareness of and
involvement in healthy eating and investigated how they related it to their roles in the
healthy eating socialisation taking place within the family. As a follow-up on a healthy
ix
eating intervention, 38 adolescents and their respective families participated in depthinterviews and a practical exercise on daily fruit and vegetable intake. Results
demonstrated that adolescents were found to adopt two strategies: a direct one placing
demands on parents or a cooperative one helping parents. Parents initiated dialogues with
family members about healthy eating and felt responsible as role models often fulfilling
the adolescents’ demands and acknowledging their help. The findings confirm that
parents still have the upper hand, when it comes to healthy eating, but with adolescents as
active players confirming the notion of consumer socialisation as bidirectional processes.
The study supplements previous research by including adolescents’ immediate family as a
unit of analysis. By taking an intra-familiar systemic approach to studying family
socialisation, this paper contributes with identifying and understanding barriers and
facilitators of adolescents’ healthy eating.
The second research question is answered in research paper 2. The paper
aimed at testing whether the common belief that children become increasingly influenced
by friends at the expense of parents during adolescence is also true for healthy eating.
Surveys were completed by 757 adolescent-parent dyads. The paper draws on Social
Cognitive Theory and The Focus Theory of Normative Conduct and finds that parents
remain the main influencer with what they do (descriptive norms) being more important
than what they say (injunctive norms). The study contributes to a more comprehensive
understanding of what influences adolescent healthy eating by comparing the influence of
entangled social spheres (parents and friends) while also controlling for personal factors
such as the adolescent’s self-efficacy and outcome expectations. No previous studies have
included all these factors in the same analysis. The implications of the study are that (1)
healthy eating interventions should aim at strengthening self-efficacy and positive
outcome expectations among adolescents, (2) the family context should be included when
x
implementing healthy eating interventions, and (3) parents’ awareness of their influence
on adolescents’ healthy eating should be reinforced.
The third research question is answered in research paper 3. The paper
investigated the effects of a healthy eating intervention by employing feedback via text
messaging during 11 weeks in order to improve adolescents’ behaviour, self-efficacy and
outcome expectations regarding fruit and vegetable intake. A pre- and post-survey was
completed by 1488 adolescents randomly allocated to a control group and two
experimental groups. Both experimental groups set weekly goals on fruit and vegetable
intake, reported their consumption daily and subsequently received feedback on their
performance via mobile text messaging (SMS). One of the experimental groups received,
in addition, a 45-minute nutrition education session from a dietician during school. The
direct effects of the interventions were not significant. However, for adolescents
participating in the SMS routines, there were significant effects of the level of activity in
the intervention, reflected in the number of sent text messages. Participants sending more
than half of the possible text messages significantly increased their fruit and vegetable
intake. Participants sending between 10 and 50% of the possible text messages
experienced a significant drop in self-efficacy, and those sending less than 10%
experienced a significant drop in outcome expectations. The findings suggest that
participants’ active engagement in an intervention is crucial to its success. This has
implications for the design and execution of health-promoting interventions.
xi
xii
TABLE OF CONTENTS
1. BACKGROUND ............................................................................................................1
1.1. Causes and consequences of not eating healthy – the special case of adolescents .......1
1.2. Research gaps and research questions ..........................................................................6
1.3. Scope of thesis ............................................................................................................11
1.4. Thesis structure ...........................................................................................................14
References .......................................................................................................................... 16
2. STATE-OF-THE-ART AND THEORETICAL FRAMEWORK ........................... 25
2.1. Consumer behaviour and behaviour change ............................................................... 25
2.2. A socio-cognitive theory approach to consumer socialisation ...................................28
2.3. A socio-cognitive theory approach to behaviour change ............................................34
2.4. Feedback interventions ............................................................................................... 37
2.5. Theoretical approach – overview ................................................................................ 41
References .......................................................................................................................... 42
3. METHODOLOGY AND RESEARCH DESIGN ..................................................... 51
3.1. Studying healthy eating............................................................................................... 51
3.2. Studying adolescents ...................................................................................................52
3.3. Mixed methods approach ............................................................................................ 54
3.4. Design of empirical studies ......................................................................................... 58
3.4.1. Study 1: Qualitative study ................................................................................................ 58
3.4.2. Study 2: Intervention study .............................................................................................. 65
3.4.2.1. Feedback intervention ......................................................................................................... 66
3.4.2.2. Pre-and post-intervention surveys ....................................................................................... 76
References .......................................................................................................................... 79
Appendixes A-E ................................................................................................................. 85
4. RELATIONS BETWEEN RESEARCH PAPERS ................................................. 101
xiii
5. FAMILY MEMBERS’ ROLES IN HEALTHY-EATING SOCIALISATION
BASED ON A HEALTHY-EATING INTERVENTION ............................................... 105
5.1. Introduction ................................................................................................................... 106
5.2. Family members’ awareness of and involvement in healthy-eating interventions ....... 108
5.3. Family members’ roles in healthy-eating socialisation ................................................. 111
5.4. Method........................................................................................................................... 113
5.4.1. Sampling ............................................................................................................................. 113
5.4.2. Interview procedures ........................................................................................................... 114
5.4.3. Background for interviews .................................................................................................. 115
5.4.4. Data analysis ....................................................................................................................... 116
5.5. Results ........................................................................................................................... 117
5.5.1. Awareness of and involvement in healthy eating based on a healthy-eating intervention.. 118
5.5.1.1. Children ...................................................................................................................... 119
5.5.1.2. Parents ........................................................................................................................ 122
5.5.2. Family member’s roles in healthy eating socialisation ....................................................... 125
5.5.2.1. Children ...................................................................................................................... 125
5.5.2.2. Parents ........................................................................................................................ 127
5.6. Discussion ..................................................................................................................... 130
Acknowledgements .............................................................................................................. 135
References ............................................................................................................................ 136
6. FOLLOWING FAMILY OR FRIENDS. SOCIAL NORMS IN ADOLESCENT
HEALTHY EATING ......................................................................................................... 141
6.1. Introduction ................................................................................................................... 142
6.2. Methods ......................................................................................................................... 146
6.2.1. Participants and procedure .................................................................................................. 146
6.2.2. Measures ............................................................................................................................. 148
6.3. Results ........................................................................................................................... 151
6.3.1. Step 1 .................................................................................................................................. 152
6.3.2. Step 2 .................................................................................................................................. 153
6.3.3. Step 3 .................................................................................................................................. 154
6.4. Discussion ..................................................................................................................... 157
Acknowledgements .............................................................................................................. 161
References ............................................................................................................................ 162
xiv
7. TEXTING YOUR WAY TO HEALTHIER EATING? EFFECTS OF
PARTICIPATING IN A FEEDBACK INTERVENTION USING TEXT
MESSAGING ON ADOLESCENTS’ FRUIT AND VEGETABLE INTAKE ........167
7.1 Introduction ................................................................................................................ 168
7.2. Hypotheses ................................................................................................................ 171
7.3. Method ...................................................................................................................... 172
7.3.1. Participants ..................................................................................................................... 172
7.3.2. Interventions ................................................................................................................... 173
7.3.3. The survey ...................................................................................................................... 175
7.3.4. Statistical analyses.......................................................................................................... 176
7.4. Results ....................................................................................................................... 177
7.5. Discussion ................................................................................................................. 183
Funding ............................................................................................................................ 190
Acknowledgements ..........................................................................................................190
References ........................................................................................................................ 191
Appendix A ...................................................................................................................... 197
8. CONCLUSION AND IMPLICATIONS .................................................................199
8.1. Specific research paper contributions .......................................................................201
8.2. Limitations ................................................................................................................ 206
8.3. Implications............................................................................................................... 208
8.3.1. Research implications .................................................................................................... 208
8.3.2. Theoretical implications ................................................................................................. 210
8.3.3. Policy implications ......................................................................................................... 213
References ........................................................................................................................ 215
xv
xvi
1. BACKGROUND
In recent decades, levels of overweight and obesity among children and
adolescents have been increasing globally and especially in economically developed
countries (Hedley et al., 2004; Lobstein & Frelut, 2003). According to the World Health
Organization (WHO), the proportion of young children who are overweight or obese
(according to the Body Mass Index) have increased about 60% in two decades (WHO,
2003) and the European Commission (2007) estimated in 2006 that 30% of European
children were overweight. Approximately one half of overweight adolescents and over
one third of overweight children remain obese as adults (Wang & Lobstein, 2006). The
prevalence of overweight children and adolescents (4-18 years) has also increased
significantly in Denmark (Matthiessen et al., 2008). By sticking to a healthy diet and
exercising, overweight and the consequential health problems can be avoided. However,
eating healthy can be a challenge in today’s society. This thesis aims at exploring how
adolescents’ healthy eating behaviour is established, influenced and can be improved.
1.1. Causes and consequences of not eating healthy – the special case of adolescents
Adolescence is one of the most dynamic and complex transitions in our life
span. WHO defines adolescence as a period from age 10 to 19 (WHO, 2014), where the
adolescent is in a transformation stage between childhood and adulthood. This period
brings about rapid growth in height and weight, sexual maturity and more advanced
cognitive development (Schickedanz, Schickedanz, Forsyth, & Forsyth, 2001). The
physical, developmental and social changes that occur during adolescence can markedly
affect eating behaviours and nutritional health (Story, 2002). Adolescence can be seen as
a window of opportunity to rectify undesirable behaviours that have been established
1
during childhood. WHO focuses on adolescents’ health with “A second chance in the
second decade” (WHO, 2014) which deals with life-threatening problems such as undernutrition and HIV, but also with the importance of healthy nutrition. Hence, adolescents
should not suffer from causes that are preventable or treatable such as overweight and
obesity.
Since the share of adolescents being overweight and not eating according to
the official dietary guidelines is increasing, the causes for this should be investigated.
Overweight and obesity result from the interaction of many factors, including genetic,
metabolic, behavioural and environmental influences. However, the rapidness of the
development suggests that the behavioural and environmental influences, rather than
biological changes, have fuelled this development (Stein & Colditz, 2004). Political and
economic developments in Western societies throughout the 20th century have changed
the way human beings live, work and eat. Urbanization, industrialization and increased
purchasing power have provided consumers with easy access to energy-dense food
(Chamberlain, 2004) while working in less physical demanding jobs (Hill & Melanson,
1999; Witkowski, 2007). Unhealthy eating and low physical activity levels are the main
contributors to overweight which appears when calorie intake exceeds calorie spending
over a longer period.
The consequences of not eating healthy can be overweight and eventually
obesity which is linked to a variety of chronic diseases such as cardiovascular disease
(Seidell, Verschuren, van Leer, & Kromhout, 1996), hypertension (Must et al., 1999),
diabetes (Stein & Colditz, 2004), different types of cancer (Calle, Rodriguez, WalkerThurmond, & Thun, 2003) and other diseases (WHO, 2003). Overweight in adolescence
is associated with an increased risk of mortality from all these diseases independent of
weight status in adulthood (Must, Jacques, Dallal, Bajema, & Dietz, 1992). Therefore, it
2
is important to prevent adolescents’ overweight. Inappropriate eating practices in
adolescence also affect the risk for eating disorders, iron deficiency, under-nutrition,
ability to concentrate and learn and thereby school performance (Story, 2002). Besides
the individual’s risk of diseases, treating overweight-related diseases in the health care
system places huge costs on society. Thus, it has been stated that “the obesity epidemic is
not only impairing individuals’ lives, but also societies’ sustainability” (Reisch &
Gwozdz, 2011), where social cohesion, equity and fairness is eroded due to the
consequences of obesity.
Eating is a vital consumer behaviour. With the birth of a child, parents or
other main caretakers feed and teach the child how and what to eat (Bauer, Laska,
Fulkerson, & Neumark-Sztainer, 2011; Birch & Fisher, 1998). Consumer socialisation
research places parents as the main agents in the primary socialisation process, where
they both explicitly and implicitly transmit consumer-related orientations to the child
(Ekström, 2006; John, 1999; Ward, 1974). In the secondary socialisation process, peers,
school and media become more influential (Dotson & Hyatt, 2005; John, 1999; Moschis,
1985). The continuous process of creating and shaping eating behaviours is important,
since eating behaviours established in childhood are often carried into adulthood (KleinHessling, Lohaus, & Ball, 2005; Knai, Pomerleau, Lock, & McKee, 2006; Stice, Shaw, &
Marti, 2006). According to Due et al. (2011) most health behaviours are socially
patterned in adolescence and track into adulthood with higher risks of adverse outcomes
among individuals from lower socioeconomic positions. There is evidence that dietary
quality declines from childhood to adolescence with decreasing intake of fruit and
vegetables and increasing soft drink consumption (Lytle, Seifert, Greenstein, &
McGovern, 2000; WHO, 2003). This can be due to growing independence and freedom to
select foods in accordance with their own individual preferences (Brown, McIlveen, &
3
Strugnell, 2000). Eating away from home, concern with physical appearance and body
weight, the need for peer acceptance and busy schedules all characterize adolescence and
can effect eating patterns and food choice (Story, 2002). This indicates that healthy eating
can be especially challenged in adolescence (Fitzgerald, Heary, Kelly, Nixon, & Shevlin,
2013).
In preventing overweight and obesity, healthy eating is (alongside physical
activity) considered a significant contributor. Promotion of healthy living to secure good
quality of life and prevent diseases has been a main target for WHO, who defines health
as “a state of complete physical, mental, and social well-being and not merely the absence
of disease or infirmity” (2011). Healthy eating refers, hence, to behaviours securing a
properly functioning body through intake of healthy food and drinks. Policy makers have
acknowledged that the responsibility of eating healthy and exercising in todays
“obesogenic” urbanised consumption environments (Lobstein, Baur, & Uauy, 2004;
Reisch & Gwozdz, 2011) cannot be placed on individuals alone. Most countries in the
economically developed part of the world therefore have national action plans for fighting
overweight and obesity, and a main component is providing consumers with information
about health. In Denmark, official recommendations regarding healthy food consumption
has been communicated to the wider public by Danish authorities for almost half a
century (Ebdrup, Persson, & Secher, 2013). In recent decades, healthy eating has been
heavily promoted in Denmark through specific campaigns targeting for instance fruit and
vegetable intake (“six-a-day” recommending a daily intake of 600 grams of fruit and
vegetables), whole grain intake (Kyrø et al., 2012) and fish intake (Scholderer & Grunert,
2001).
Information campaigns on healthy eating aim at providing consumers with
information about types and amounts of food most favourable for health. They are most
4
often evaluated in terms of their impact on consumers’ nutritional knowledge and/or
behaviour. However, health information campaigns may not always have a positive
effect, since, as Wardle and Huon points out, such campaigns are often based on “the
assumption that the rational consumer will, other things being equal, choose the food that
they know is healthier” (2000, p. 39). However, this is not always the case – especially
not with children and adolescents being in a more “irrational” life phase. Some studies
have found adults’ nutritional knowledge to be significantly related to healthy eating (see
for instance Wardle, Parmenter, & Waller, 2000), while others find that nutritional
knowledge is a necessary, but not a sufficient factor for changing dietary behaviour
(Sichert-Hellert et al., 2011; Worsley, 2002). When it comes to healthy eating and food
choice, nutritional knowledge competes with environmental factors such as availability
and social norms regarding healthy eating (Sichert-Hellert et al., 2011), but also with
intra-personal factors such as perceived healthfulness (Davenport, Radcliffe, Chen, &
Cullen, 2014), personal motivation (Worsley, 2002), taste preferences (Wardle & Huon,
2000), familiarity/habit and ability of self-regulation (Contento, Williams, Michela, &
Franklin, 2006), Hence, it is not unexpected that several studies on adolescents’ healthy
eating point to a knowledge-behaviour gap, where adolescents demonstrate knowledge
regarding healthy foods and healthy eating, but rarely puts it into practice (Bech-Larsen &
Kazbare, 2014; Croll, Neumark-Sztainer, & Story, 2001; Wiggins, 2004). It has been
found that communication strategies should not be exclusively based on providing
adolescents with more information on a proper diet (Pich, Ballester, Thomàs, Canals, &
Tur, 2011), but should emphasise the fun and enjoyable attributes of healthy eating (Chan
& Tsang, 2011). However, it is also important to look into the more structural barriers for
adolescents’ healthy eating such as limited availability, lack of concern and time (Croll et
al., 2001), relative cheapness and personal taste preference for fast food (Shepherd &
5
Raats, 2006), cravings for unhealthy food and not being too “overly healthy” to important
others (Bech-Larsen & Kazbare, 2014). Parents’ educational background and their
nutritional knowledge have also proved to be important (Jenkins & Horner, 2005; SichertHellert et al., 2011). Facilitators for healthy eating include support from family (Chan,
Prendergast, Grønhøj, & Bech-Larsen, 2010), wider availability of healthy foods, desire
to look after one’s appearance and will-power (Shepherd et al., 2006). When promoting
healthy eating to adolescents, it is therefore essential to understand the complexity of
causes for not eating healthy and identify and deal with both barriers and facilitators for
healthy eating.
1.2. Research gaps and research questions
Research on eating behaviours has focused extensively on children’s eating
behaviours in relation to development of food preferences (e.g. Birch & Fisher, 1998;
Wardle & Huon, 2000), parents’ feeding practices (e.g. Melbye, Øgaard, & Øverby,
2013; Wardle & Carnell, 2007), food choice (e.g. Douglas, 1998) and food intake (e.g.
Muñoz, Krebs-Smith, Ballard-Barbash, & Cleveland, 1997). Young people under the age
of 18 are often defined as children (France, 2004), and studies on children’s eating
behaviour have therefore often included adolescents as well. In a historical context
adolescents have always been of interest to researchers and policy makers because of
their “problematic” nature (France, 2004). In recent decades, where eating behaviours
have become more problematic causing rising levels of overweight among children and
adolescents and impacting their entire life course, researchers have come to focus more
on adolescents’ eating behaviours as well (e.g. Savage, Fisher, & Birch, 2007),
acknowledging that adolescents are important to study in their own right. Studies of
adolescents’ eating behaviours have looked into eating disorders (see for instance
6
Herpertz-Dahlmann, Wille, Hölling, Vloet, & Ravens-Sieberer, 2008), body image
concerns (Littleton & Ollendick, 2003), dieting (Hill, Oliver, & Rogers, 1992; NeumarkSztainer et al., 2006) and treatment of overweight and obesity (see Stice et al., 2006, for a
review). However, it is important to recognise that healthy eating is first and foremost an
eating behaviour (Kazbare, 2010) – and not a problematic behaviour from the outset. It
cannot be isolated from other aspects of daily life, and research on healthy eating should
therefore take both the context (socio-structural/environmental factors) and the
individual’s motivation for eating healthy into account.
The context for most adolescents is the family-setting, since they still live at
home. However, the family as a consumer entity is underrepresented in consumer
behaviour research (Grønhøj, 2002; Nørgaard, 2009). Consumer socialisation research
provides a basis for studying the interaction between adolescent and family regarding
healthy eating. By comparing adolescents’ and one or more family members’ (often
parents’) attitudes/behaviours/norms, studies have been able to determine the character of
the interaction and the direction and level of influence. This has been done in the area of
family decision-making, where Foxman et al. (1989) found that children had some
influence in purchase decisions for a variety of products (no food products). Nørgaard’s
(2009) family decision-making study of food buying found that children can influence
parents directly through demands and indirectly through suggestions of healthy food
products. However, while acknowledging the importance of family decision-making as an
aspect of consumer socialisation, studies with a wider perspective on healthy eating
socialisation than food buying are few. Hence, the area of family interaction and family
members’ roles regarding healthy eating socialisation is still underexposed. This leads to
the first research question (explored in research paper 1 in Chapter 5):
1. Which roles do parents and adolescents have in healthy eating socialisation?
7
Consumer socialisation research regards parents as primary socialisation
agents for children and adolescents’ healthy eating (Birch & Fisher, 1998; Savage et al.,
2007). Yet, with the growing independence and more time spent away from home in
adolescence, the influence of especially friends as secondary socialisation agents cannot
be ignored. A review of 58 studies on environmental factors (understood as “anything
outside the individual”) influencing adolescents’ eating behaviours shows that parents’
food intake is most consistently associated with adolescents’ intake (van der Horst et al.,
2007). However, most of these studies focus on the household level thereby limiting the
room for friends’ influence. Influence of parents and friends is mostly assessed
separately, and only few studies have compared the social influence of parents and friends
on adolescent healthy eating (see for instance Salvy, Elmo, Nitecki, Kluczynski, &
Roemmich, 2011). John (1999) acknowledges that the understanding of the social
environment in consumer socialisation research is insufficient. Especially, there are gaps
in the conceptualisation of entangled social spheres (i.e. family, friends, media, school)
directly and indirectly influencing the adolescent. However, the adolescent should not be
seen as a passive receiver of socialisation efforts. Personal factors (i.e. affective, cognitive
factors) influencing healthy eating should also be taken into account, and according to
social cognitive theory, self-efficacy and outcome expectations are important personal
factors in self-regulation (Bandura, 1997). Acknowledging the importance of both
environmental and personal factors for healthy eating, the second research question
(explored in research paper 2, Chapter 6) is:
2. How does the social influence from parents and friends compared to personal factors
impact adolescents’ healthy eating?
Acknowledging that healthy eating is an eating behaviour taking place every
day in various contexts, it is also relevant to explore how adolescents’ personal
8
motivation for behaviour change can be influenced. WHO recommends public health
policies to focus more on prevention of health-compromising behaviours (WHO, 2014).
Promoting healthy eating behaviours is crucial when preventing overweight. However,
despite government initiated health campaigns and official dietary recommendations,
adolescents do not eat in accordance with the recommendations (Krølner et al., 2011;
Lynch et al., 2014; Rasmussen et al., 2006; WHO, 2003). The fact that most information
campaigns on healthy eating have concentrated on the adult consumer (Brown et al.,
2000) might play a role here. Nevertheless, other initiatives such as behaviour change
interventions targeting adolescents’ healthy eating have been implemented. The literature
is rich on clinical interventions with overweight or obese adolescents (e.g. Eliakim et al.,
2002), but also community-based interventions with adolescents (not necessarily
overweight) in a school setting (see for instance Anderson et al., 2006; Gorely, Nevill,
Morris, Stensel, & Nevill, 2009) or in a family setting (e.g. Pearson, Atkin, Biddle, &
Gorely, 2010) are studied. How to obtain positive effects in terms of positive behaviour
changes is the main challenge for intervention researchers, since obesity prevention
programmes tend to produce mixed and modest results (Branscum, Sharma, Wang,
Wilson, & Rojas-Guyler, 2013). Associations between significant effects and the design
and implementation of interventions are very hard to point out, since the research on this
is fragmented and still in an early stage (Hardeman et al., 2005; Stice et al., 2006). Michie
et al. (2008) underline the importance of a theoretical basis when designing interventions.
A meta-analytic review of 64 obesity prevention programmes for children and
adolescents by Stice et al. (2006) provides a detailed account of what makes interventions
work: targeting adolescents with brief, pilot-tested programmes focusing on one health
behaviour with an opt-in approach for participation. These findings can enrich a process
evaluation of an intervention which is important when no significant results of the
9
intervention are found. For instance, Bere et al. (2006) found no results of their fruit and
vegetable intervention although a process evaluation found that the intervention was
implemented and received as intended. They concluded that the intervention did not
succeed in changing the strongest correlates of intake which were preferences for and the
accessibility of fruit and vegetables.
For interventions to be attractive for health promotors, they should
document results, be cost-effective, be easy to implement and be able to document the
behaviour change process based on a solid theoretical framework. Adolescents can be a
difficult target group for behaviour change efforts, and promotion of healthy eating
should be fun and appealing (Chan & Tsang, 2011). Some interventions targeting
adolescents have therefore used the technological platform of mobile phones’ Short
Messaging System (SMS) which is one of adolescents’ preferred communication tools
(see section 2.4). However, it is important to underline that it is more likely the theoretical
framework and careful design of the intervention that can facilitate behaviour change –
rather than the technological platform in itself. The third research question (explored in
research paper 3, Chapter 7) is therefore:
3. Can a feedback intervention based on socio-cognitive theory and using text messaging
improve adolescent healthy eating?
This thesis takes a broader perspective on promoting healthy eating to
adolescents, when trying to understand the origin of eating behaviours, the factors
influencing healthy eating and the adolescents’ motivation for behaviour change. By
answering the three research questions, the overall aim of this thesis is to provide insights
into adolescent healthy eating by using a socio-cognitive approach on consumer
socialisation and behaviour change as input to understand and possibly improve
adolescents’ eating behaviours.
10
1.3. Scope of thesis
This thesis focuses on healthy eating. Healthy eating is important for a
proper function of our bodies. However, it is recognised that living a healthy life consists
of more factors than just healthy eating such as physical activity, good mental health,
social relations etc. Healthy eating is intake that fulfils various nutritional needs, such as
fruit and vegetables, fish, fibre, vitamins etc. This thesis focuses specifically on the intake
of fruit and vegetables. There are several reasons for this: First of all, eating more fruit
and vegetables (than what is now the norm in countries like Denmark) has been singled
out as one of the most important components of a healthy diet (Kazbare, 2010; Thomas et
al., 2003). Specifically, most Danish adolescents do not eat the recommended daily
amount of 600 grams fruit and vegetables. During the last 25 years, the percentage of
adolescents eating fruit daily has decreased for both 11 and 15 year-old girls and boys
(e.g., for 15 year-old boys from 77% to 51 %), and the same for vegetables (e.g., for 15
year-old girls from 69 % to 63 %). In the same period, campaign activities for increasing
fruit and vegetable intake have increased (Rasmussen, Pedersen, & Due, 2015). Second,
studies of adolescents’ perception of healthy eating have found that fruit and vegetables
were the most commonly mentioned healthy foods (e.g. Croll et al., 2001; McKinley et
al., 2005; O'Dea, 2003). These findings support this thesis’ assumption that an increase in
fruit and vegetable intake should be a key element in promoting healthy eating among
adolescents.
Third, there are practical reasons for this focus. Fruit and vegetables are
considered easier for adolescents to count in an intervention (such as the one described in
this thesis) than for instance intake of wholegrain or fish. Even though a review found
that fruit and vegetables were perceived by adolescents as inconvenient snack food since
they are not instantly available and need to be washed, dried, peeled or cooked before
11
consumption (Krølner et al., 2011), fruit and vegetables are still considered easier to
access for adolescents (requiring only limited cooking skills) than other healthy foods
(e.g., fish or lean meat). In Denmark, vegetables are mainly eaten for lunch (as part of a
homemade lunch box) and dinner, while fruits are often eaten as snacks in-between meals
or as a “dessert” after lunch (Christensen, Kørup, Trolle, Matthiessen, & Fagt, 2012).
These foods are hence not only eaten in the company of parents and siblings, but also
together with friends (e.g., at school and after school) who are found to influence
adolescents’ eating behaviours to some degree (Bruening et al., 2012; Contento et al.,
2006; Howland, Hunger, & Mann, 2012; Nørgaard, Hansen, & Grunert, 2013; Salvy et
al., 2011).
Still, parents are the main food providers (Lau, Quadrel, & Hartman, 1990;
Savage et al., 2007) and availability is a significant predictor of fruit and vegetable intake
among adolescents (Contento et al., 2006; Cullen et al., 2003). In this study it is not,
however, measured to which extent parents actually provide fruit and vegetables
compared to other healthy and unhealthy foods. Hence, this thesis does not cover all
aspects of healthy eating. It is for instance acknowledged that excessive intake of
especially fruit can lead to too high calorie intake. Hence, in order to obtain a complete
picture of healthiness in terms of eating the quantity of foods consumed over a longer
period of time, and the total energy-balance, would need to be assessed, which it is not in
this thesis. It is also acknowledged that increased fruit and vegetable intake does not
necessarily mean a decrease in unhealthy food intake – and that reducing unhealthy eating
is an important topic in its own right, also not addressed in this thesis.
It is also acknowledged that a significant part of daily eating behaviours
consists of habits (Riet, Sijtsema, Dagevos, & De Bruijn, 2011) which “are a form of
automaticity in responding that develops as people repeat actions in stable circumstances”
12
(Verplanken, 2006; Verplanken & Wood, 2006). By studying adolescents’ healthy eating
in a family context and exploring the social influence of especially parents and friends,
the importance of the environment is acknowledged. However, this thesis does not look
into the environmental cues or other factors triggering the automaticity and efficiency
characterising habitual behaviour. Furthermore, it may be surmised that children and
adolescents are in a process of establishing habits, but that they do not yet possess as
strongly established habits as adults. Hence, while studying healthy eating as habitual
behaviour could be valuable in this age group, priority is here given to treating healthy
eating behaviour as a result of socio-cognitive processes.
The focus of this thesis is on two target groups: younger (11-12 years) and
older (15-16 years) adolescents. These two groups are chosen because they possess
different cognitive and social competences, but also because they are at different places in
the transition from child to adult. Most adolescents still live at home and are – according
to consumer socialisation theory – heavily influenced by their parents, friends, school,
media etc. It is acknowledged that it is important to study adolescents in their social
context. However, in this thesis, the focus is mainly on adolescents, their parents (no
particular differentiation between father and mother or other family members such as
siblings or step-parents) and to a lesser extent friends and not school, media/social media.
By adopting a theoretical approach that views human agency as operating within an
interdependent causal structure, it is acknowledged that adolescents are influenced not
only by family and friends, but also by their broader social, physical and cultural
environment.
13
1.4. Thesis structure
This dissertation consists of eight chapters. As already presented, chapter 1
introduces adolescents’ healthy eating as a subject for research, the research gaps and the
specific research questions this thesis aims to answer. Chapter 2 introduces the theoretical
framework of social cognitive theory in relation to consumer socialisation and behaviour
change. The chapter also reviews previous studies on adolescents in relation to healthy
eating, consumer socialisation, social influence and behaviour change interventions using
feedback. Chapter 3 presents methodological reflections on the study object (healthy
eating and adolescents) and the nature of the study (mixed method approach). Then the
research methods and designs of the two studies developed and used in this thesis are
presented. Chapter 4 gives a summary of the three research papers and reflects on the
relations between them. The three research papers are presented in chapters 5, 6 and 7.
They are reported in full length in their original published or submitted versions. Chapter
8 presents the concluding remarks of the dissertation, the specific research contributions,
limitations, implications and suggestions for future research.
14
15
References
Anderson, A. S., Porteous, L. E. G., Foster, E., Higgins, C., Stead, M., Hetherington, M., .
. . Adamson, A. J. (2006). The impact of a school-based nutrition education
intervention on dietary intake and cognitive and attitudinal variables relating to
fruits and vegetables. Public Health Nutrition, 8(6), 650-656.
Bandura, Albert. (1997). Self-efficacy: The exercise of control. New York: W. H.
Freeman.
Bauer, K W, Laska, Melissa N, Fulkerson, Jayne A, & Neumark-Sztainer, D. (2011).
Longitudinal and secular trends in parental encouragement for healthy eating,
physical activity and dieting throughout the adolescent years. Journal of
Adolescent Health, 49(3), 306. doi: 10.1016/j.jadohealth.2010.12.023
Bech-Larsen, Tino, & Kazbare, Laura. (2014). Perceptions of healthy eating in
transitional phases of life. British Food Journal, 116(4), 570-584. doi:
doi:10.1108/BFJ-05-2012-0117
Birch, L., & Fisher, J. (1998). Development of eating behaviors among children and
adolescents. Pediatrics, 101(3), 539-549.
Branscum, Paul, Sharma, Manoj, Wang, Lihshing Leigh, Wilson, Bradley, & RojasGuyler, Liliana. (2013). A process evaluation of a social cognitive theory–based
childhood obesity prevention intervention: The comics for health program. Health
Promotion Practice, 14(2), 189-198. doi: 10.1177/1524839912437790
Brown, Karen, McIlveen, Heather, & Strugnell, Christopher. (2000). Nutritional
awareness and food preferences of young consumers. Nutrition & Food Science,
30(5), 230-235. doi: 10.1108/00346650010340963
Bruening, M. , Eisenberg, M. , MacLehose, R. , Nanney, M. , Story, M. , & NeumarkSztainer, D. (2012). Relationship between adolescents’ and their friends’ eating
behaviors: Breakfast, fruit, vegetable, whole-grain and dairy intake. Journal of
Academy of Nutrition and Dietetics, 112, 1608-1613. doi:
10.1016/j.jand.2012.07.008
Calle, Eugenia E. PhD, Rodriguez, Carmen M. D. M. P. H., Walker-Thurmond, Kimberly
B. A., & Thun, Michael J. M. D. (2003). Overweight, obesity and mortality from
cancer in a prospectively studied cohort of U.S. adults. The New England Journal
of Medicine, 348(17), 1625-1638.
Chamberlain, Kerry. (2004). Food and health: Expanding the agenda for health
psychology. Journal of Health Psychology, 9(4), 467-481. doi:
10.1177/1359105304044030
16
Chan, Kara, Prendergast, Gerard, Grønhøj, Alice, & Bech-Larsen, Tino. (2010). The role
of socializing agents in communicating healthy eating to adolescents: A cross
cultural study. Journal of International Consumer Marketing, 23(1), 59–74. doi:
10.1080/08961530.2011.524578
Chan, Kara, & Tsang, Lennon. (2011). Promote healthy eating among adolescents: A
Hong Kong study. The Journal of Consumer Marketing, 28(5), 354-362. doi:
10.1108/07363761111150008
Christensen, L. M., Kørup, K., Trolle, E., Matthiessen, J., & Fagt, S. (2012). Børn og
unges måltidsvaner 2005-2008 [Children and adolescents' meal habits 2005-2008]
(1 ed.): DTU Food, Department of Nutrition
Contento, Isobel R., Williams, Sunyna S., Michela, John L., & Franklin, Amie B. (2006).
Understanding the food choice process of adolescents in the context of family and
friends. Journal of Adolescent Health, 38(5), 575-582. doi:
10.1016/j.jadohealth.2005.05.025
Croll, J. K. , Neumark-Sztainer, D., & Story, M. (2001). Healthy eating. What does it
mean to adolescents? Journal of Nutrition Education (33), 193-198.
Cullen, Karen Weber, Baranowski, Tom, Owens, Emiel, Marsh, Tara, Rittenberry,
Latroy, & de Moor, Carl. (2003). Availability, accessibility, and preferences for
fruit, 100% fruit juice, and vegetables influence children's dietary behavior.
Health Education & Behavior, 30(5), 615-626. doi: 10.1177/1090198103257254
Davenport, L. A. , Radcliffe, J. , Chen, T.-A. , & Cullen, Karen. (2014). Adolescents who
perceive their diet as healthy consume more fruits, vegetables and milk and fewer
sweet drinks. International Journal of Child Health and Nutrition, 3, 124-129.
doi: 10.6000/1929-4247.2014.03.03.2
Dotson, Michael J., & Hyatt, Eva M. (2005). Major influence factors in children's
consumer socialization. Journal of Consumer Marketing, 22(1), 35-42.
Douglas, Lesley. (1998). Children’s food choice. Nutrition & Food Science, 98(1), 14-18.
doi: 10.1108/00346659810196273
Due, Pernille, Krølner, Rikke, Rasmussen, Mette, Andersen, Anette, Damsgaard, Mogens
Trab, Graham, Hilary, & Holstein, Bjørn E. (2011). Pathways and mechanisms in
adolescence contribute to adult health inequalities. Scandinavian Journal of
Public Health, 39(6), 62-78. doi: 10.1177/1403494810395989
Ebdrup, Niels, Persson, Charlotte Price, & Secher, Kristian. (2013). Sære kostråd har
præget Danmark i århundreder [Weird dietary advises have shaped Denmark for
centuries]. www.videnskab.dk. Retrieved 18th of January, 2015, from
http://videnskab.dk/kultur-samfund/saere-kostrad-har-praeget-danmark-iarhundreder
17
Ekström, Karin M. . (2006). Consumer socialization revisited. In R. Belk (Ed.), Research
in Consumer Behavior (pp. 71-98): Emerald Group Publishing Limited.
Eliakim, Alon, Kaven, Galit, Berger, Isaac, Friedland, Orit, Wolach, Baruch, & Nemet,
Dan. (2002). The effect of a combined intervention on body mass index and
fitness in obese children and adolescents - a clinical experience. European Journal
of Pediatrics, 161(8), 449-454. doi: 10.1007/s00431-002-0980-2
EU-Commission. (2007). A strategy for Europe on nutrition, overweight and obesity
related health issues. White paper. Retrieved 14th of July, 2014, from URL
http://ec.europa.eu/health/ph_determinants/life_style/nutrition/documents/nutritio
n_wp_en.pdf
Fitzgerald, A., Heary, C., Kelly, C., Nixon, E., & Shevlin, M. (2013). Self-efficacy for
healthy eating and peer support for unhealthy eating are associated with
adolescents’ food intake patterns. Appetite, 63, 48-58. doi:
10.1016/j.appet.2012.12.011
Foxman, Ellen R., Tansuhaj, Patriya S., & Ekström, Karin M. (1989). Family members'
perceptions of adolescents' influence in family decision making. Journal of
Consumer Research, 15(4), 482-491.
France, Alan. (2004). Young people. In S. Fraser, V. Lewis, S. Ding, M. Kellett & C.
Robinson (Eds.), Doing research with children and young people (pp. 175-190).
London: SAGE Publications.
Gorely, Trish, Nevill, Mary E, Morris, John G, Stensel, David J, & Nevill, Alan. (2009).
Effect of a school-based intervention to promote healthy lifestyles in 7-11 year old
children. The International Journal of Behavioral Nutrition and Physical Activity,
6(1), 5-16. doi: 10.1186/1479-5868-6-5
Grønhøj, Alice. (2002). Miljøvenlig adfærd i familien: Et studie af familiemedlemmernes
involvering og sociale interaktion [Pro-environmental behaviour in the family: A
study of family members' involvement and social interaction]. Århus: Institut for
Markedsøkonomi.
Hardeman, Wendy, Sutton, Stephen, Griffin, Simon, Johnston, Marie, White, Anthony,
Wareham, Nicholas J., & Kinmonth, Ann Louise. (2005). A causal modelling
approach to the development of theory-based behaviour change programmes for
trial evaluation. Health Education Research, 20(6), 676-687. doi:
10.1093/her/cyh022
Hedley, A. A., Ogden, C. L., Johnson, C. L., Carroll, M. D., Curtin, L. R., & Flegal, K.
M. (2004). Prevalence of overweight and obesity among us children, adolescents
and adults, 1999-2002. JAMA, 291(23), 2847-2850. doi:
10.1001/jama.291.23.2847
18
Herpertz-Dahlmann, Beate, Wille, Nora, Hölling, Heike, Vloet, TimoD, & RavensSieberer, Ulrike. (2008). Disordered eating behaviour and attitudes, associated
psychopathology and health-related quality of life: Results of the BELLA study.
European Child & Adolescent Psychiatry, 17(1), 82-91. doi: 10.1007/s00787-0081009-9
Hill, A. J., Oliver, S., & Rogers, P. J. (1992). Eating in the adult world: The rise of dieting
in childhood and adolescence. British Journal of Clinical Psychology, 31(1), 95106.
Hill, J. O., & Melanson, E L. (1999). Overview of the determinants of overweight and
obesity: Current evidence and research issues. Medicine and science in sports and
exercise, 31(11), S515.
Howland, M. , Hunger, J. , & Mann, T. (2012). Friends don’t let friends eat cookies:
Effects of restrictive eating norms on consumption among friends. Appetite, 59,
505-509. doi: 10.1016/j.appet.2012.06.020
Jenkins, S., & Horner, S. D. (2005). Barriers that influence eating behaviors in
adolescents. Journal of Pediatric Nursing, 20, 258-267.
John, Deborah Roedder. (1999). Consumer socialization of children: A retrospective look
at twenty-five years of research. Journal of Consumer Research, 26(3), 183-213.
Kazbare, L. (2010). Healthy eating behaviour - a social marketing perspective. (PhD
Thesis), MAPP Centre, Aarhus School of Business, Aarhus University. Retrieved
from https://pure.au.dk/ws/files/14287/Kazbare_2010
Klein-Hessling, J., Lohaus, A., & Ball, J. (2005). Psychological predictors of healthrelated behaviour in children. Psychology, Health & Medicine, 41, 31-43.
Knai, C., Pomerleau, J., Lock, K., & McKee, M. (2006). Getting children to eat more fruit
and vegetables: A systematic review. Preventive Medicine, 42, 85-95.
Krølner, Rikke, Rasmussen, Mette, Brug, Johannes, Klepp, Knut-Inge, Wind, Marianne,
& Due, Pernille. (2011). Determinants of fruit and vegetable consumption among
children and adolescents: A review of the literature. Part II: qualitative studies. Int
J Behav Nutr Phys Act, 8(1), 112.
Kyrø, Cecilie, Skeie, Guri, Dragsted, Lars O., Christensen, Jane, Overvad, Kim,
Hallmans, Göran, . . . Olsen, Anja. (2012). Intake of whole grain in Scandinavia:
Intake, sources and compliance with new national recommendations.
Scandinavian Journal of Public Health, 40(1), 76-84. doi:
10.1177/1403494811421057
19
Lau, Richard R., Quadrel, Marilyn Jacobs, & Hartman, Karen A. (1990). Development
and change of young adults' preventive health beliefs and behavior: Influence
from parents and peers. Journal of Health and Social Behavior, 31(3), 240-259.
Littleton, Heather L., & Ollendick, Thomas. (2003). Negative body image and disordered
eating behavior in children and adolescents: What places youth at risk and how
can these problems be prevented? Clinical Child and Family Psychology Review,
6(1), 51-66. doi: 10.1023/a:1022266017046
Lobstein, T., Baur, L., & Uauy, R. (2004). Obesity in children and young people: A crisis
in public health. Obesity Revviews, 5(suppl), 4-104.
Lobstein, T., & Frelut, M. L. (2003). Prevalence of overweight among children in Europe.
Obesity Reviews, 4(4), 195-200. doi: 10.1046/j.1467-789X.2003.00116.x
Lynch, Christel, Kristjansdottir, Asa Gudrun, te Velde, Saskia J, Lien, Nanna, Roos, Eva,
Thorsdottir, Inga, . . . Yngve, Agneta. (2014). Fruit and vegetable consumption in
a sample of 11-year-old children in ten European countries – the PRO GREENS
cross-sectional survey. Public Health Nutrition, 17(11), 2436-2444. doi:
10.1017/S1368980014001347
Lytle, Leslie A., Seifert, Sara, Greenstein, Jessica, & McGovern, Paul. (2000). How do
children's eating patterns and food choices change over time? Results from a
cohort study. American Journal of Health Promotion, 14(4), 222-228. doi:
10.4278/0890-1171-14.4.222
Matthiessen, J., Groth, M. V., Fagt, S., Jensen, A.B., Stockmarr, A., Andersen, J.S., &
Trolle, E. (2008). Prevalence and trends in overweight and obesity among children
and adolescents in Denmark. Scandinavian Journal of Public Health, 36, 153-160.
McKinley, M. C., Lowis, C., Robson, P. J., Wallace, J. M. W., Morrissey, M., Moran, A.,
& Livingstone, M. B. E. (2005). It's good to talk: children's views on food and
nutrition. Eur J Clin Nutr, 59(4), 542-551.
Melbye, Elisabeth L, Øgaard, Torvald, & Øverby, Nina C. (2013). Associations between
parental feeding practices and child vegetable consumption. Mediation by child
cognitions? Appetite, 69, 23-30. doi: 10.1016/j.appet.2013.05.005
Michie, Susan, Johnston, Marie, Francis, Jill, Hardeman, Wendy, & Eccles, Martin.
(2008). From theory to intervention: Mapping theoretically derived behavioural
determinants to behaviour change techniques. Applied Psychology, 57(4), 660680. doi: 10.1111/j.1464-0597.2008.00341.x
Moschis, G. (1985). The role of family communication in consumer socialization of
children and adolescents. Journal of Consumer Research, 11(4), 898-913.
20
Muñoz, Kathryn A., Krebs-Smith, Susan M., Ballard-Barbash, Rachel, & Cleveland,
Linda E. (1997). Food intakes of US children and adolescents compared with
recommendations. Pediatrics, 100(3), 323-329. doi: 10.1542/peds.100.3.323
Must, A., Jacques, P. F., Dallal, G. E., Bajema, C.J., & Dietz, W. H. M. D. (1992). Longterm morbidity and mortality of overweight adolescents: A follow-up of the
Harvard growth study of 1922 to 1935. The New England Journal of Medicine,
327(19), 1350-1355.
Must, A., Spadano, J., Coakley, E. H., Field, A. E., Colditz, G., & Dietz, W. H. (1999).
The disease burden associated with overweight and obesity. JAMA, 282(16),
1523-1529. doi: 10.1001/jama.282.16.1523
Neumark-Sztainer, Dianne, Wall, Melanie, Guo, Jia, Story, Mary, Haines, Jess, &
Eisenberg, Marla. (2006). Obesity, disordered eating and eating disorders in a
longitudinal study of adolescents: How do dieters fare 5 years later? Journal of the
American Dietetic Association, 106(4), 559-568. doi: 10.1016/j.jada.2006.01.003
Nørgaard, Maria Kümpel. (2009). Family decision-making during food buying. (PhD
thesis), MAPP Centre, Aarhus School of Business, Aarhus University.
Nørgaard, Maria Kümpel, Hansen, Kathrine Nørgaard, & Grunert, Klaus G. (2013). Peer
influence on adolescent snacking. Journal of Social Marketing, 3(2), 176-194. doi:
10.1108/jsocm-06-2012-0028
O'Dea, Jennifer A. (2003). Why do kids eat healthful food? Perceived benefits of and
barriers to healthful eating and physical activity among children and adolescents.
Journal of the American Dietetic Association, 103(4), 497-501. doi:
http://dx.doi.org/10.1016/S0002-8223(03)00013-0
Pearson, Natalie, Atkin, Andrew J , Biddle, Stuart JH , & Gorely, Trish (2010). A familybased intervention to increase fruit and vegetable consumption in adolescents: A
pilot study. Public Health Nutrition 13 (6), 876-885.
Pich, Jordi, Ballester, Lluís, Thomàs, Mònica, Canals, Ramon, & Tur, Josep A. (2011).
Assimilating and following through with nutritional recommendations by
adolescents. Health Education Journal, 70(4), 435-445. doi:
10.1177/0017896910379695
Rasmussen, Mette, Krolner, Rikke, Klepp, Knut-Inge, Lytle, L, Brug, Johannes, Bere, E,
& Due, Pernille. (2006). Determinants of fruit and vegetable consumption among
children and adolescents: A review of the literature. Part I: quantitative studies. Int
J Behav Nutr Phys Act, 3, 22.
Rasmussen, Mette, Pedersen, Trine, & Due, Pernille. (2015). Skolebørnsundersøgelsen
2014 [Health behaviour in school-aged children (HBSC) study 2014].
21
Copenhagen: Statens Institut for Folkesundhed [The Danish National Institute of
Public Health].
Reisch, L., & Gwozdz, W. (2011). Chubby cheeks and climate change: Childhood obesity
as a sustainable development issue. International Journal of Consumer Studies,
35(1), 3-9.
Riet, van 't, J.P, Sijtsema, S.J, Dagevos, H. , & De Bruijn, G.-J. (2011). The importance
of habits in eating behaviour: An overview and recommendations for future
research. Appetite, 57(3), 585-596. doi: 10.1016/j.appet.2011.07.010
Salvy, Sarah-Jeanne, Elmo, Alison, Nitecki, Lauren A, Kluczynski, Melissa A, &
Roemmich, James N. (2011). Influence of parents and friends on children's and
adolescents' food intake and food selection. The American Journal of Clinical
Nutrition, 93(1), 87-92. doi: 10.3945/ajcn.110.002097
Savage, Jennifer S., Fisher, Jennifer Orlet, & Birch, Leann L. (2007). Parental influence
on eating behavior: Conception to adolescence. The Journal of Law, Medicine &
Ethics, 35(1), 22-34. doi: 10.1111/j.1748-720X.2007.00111.x
Schickedanz, J., Schickedanz, D., Forsyth, P. , & Forsyth, G. . (2001). Understanding
children and adolescents (4th ed.). Boston: Allyn & Bacon.
Scholderer, Joachim, & Grunert, Klaus G. (2001). Does generic advertising work? A
systematic evaluation of the Danish campaign for fresh fish. Aquaculture
Economics and Management, 5(5/6), 253-272.
Seidell, J. C., Verschuren, W., van Leer, E. M., & Kromhout, D. (1996). Overweight,
underweight and mortality: A prospective study of 48287 men and women.
Archives of Internal Medicine, 156(9), 958-963. doi:
10.1001/archinte.1996.00440090054006
Shepherd, J, Harden, A, Rees, R, Brunton, G, Garcia, J, Oliver, S, & Oakley, A (2006).
Young people and healthy eating: A systematic review of research on barriers and
facilitators. Health Education Research, 21(2), 239-257.
Shepherd, Richard, & Raats, Monique. (2006). The psychology of food choice.
Cambridge, MA: CABI Pub.
Sichert-Hellert, Wolfgang, Beghin, Laurent, De Henauw, Stefaan, Grammatikaki,
Evangelia, Hallström, Lena, Manios, Yannis, . . . Kersting, Mathilde. (2011).
Nutritional knowledge in European adolescents: Results from the HELENA
(Healthy Lifestyle in Europe by Nutrition in Adolescence) study. Public Health
Nutrition, 14(12), 2083-2091. doi: 10.1017/S1368980011001352
22
Stein, Cynthia J, & Colditz, Graham A. (2004). The epidemic of obesity. The Journal of
Clinical Endocrinology and Metabolism, 89(6), 2522-2525. doi: 10.1210/jc.20040288
Stice, Eric, Shaw, Heather, & Marti, C. Nathan. (2006). A meta-analytic review of
obesity prevention programs for children and adolescents: The skinny on
interventions that work. Psychological Bulletin, 132(5), 667-692.
Story, M, Neumark-Sztainer, D, French, S. . (2002). Individual and environmental
influences on adolescent eating behaviours. Supplement to the Journal of the
American Dietetic Association, 102(3), 40-50.
Thomas, J., Sutcliffe, K., Harden, A., Oakley, A., Oliver, S., Rees, R., . . . Kavanagh, J.
(2003). Children and healthy eating: A systematic review of barriers and
facilitators. London: EPPI-Centre, Social Science Research Unit, Institute of
Education, University of London.
van der Horst, Klazine, Oenema, A, Ferreira, I, Wendel-Vos, W, Giskes, K, van Lenthe,
F, & Brug, J. (2007). A systematic review of environmental correlates of obesityrelated dietary behaviors in youth. Health Education Research, 22(2), 203-226.
doi: 10.1093/her/cyl069
Verplanken, Bas. (2006). Beyond frequency: Habit as mental construct. British Journal of
Social Psychology, 45(3), 639-656. doi: 10.1348/014466605x49122
Verplanken, Bas, & Wood, Wendy. (2006). Interventions to break and create consumer
habits. JPP&M, 25(1), 90-103. doi: 10.1509/jppm.25.1.90
Wang, Youfa, & Lobstein, T. I. M. (2006). Worldwide trends in childhood overweight
and obesity. International Journal of Pediatric Obesity, 1(1), 11-25. doi:
10.1080/17477160600586747
Ward, Scott. (1974). Consumer socialization. Journal of Consumer Research, 1(2), 1-17.
Wardle, Jane, & Carnell, Susan. (2007). Parental feeding practices and children's weight.
Acta Pædiatrica, 96, 5-11. doi: 10.1111/j.1651-2227.2007.00163.x
Wardle, Jane, & Huon, Gail. (2000). An experimental investigation of the influence of
health information on children's taste preferences. Health Education Research,
15(1), 39-44.
Wardle, Jane, Parmenter, K., & Waller, J. (2000). Nutrition knowledge and food intake.
Appetite, 34, 269-275.
WHO. (2003). Diet, nutrition and the prevention of chronic diseases. Geneva: WHO
23
WHO. (2011). What is the definition of health? Retrieved January 27, 2015, from
http://www.who.int/suggestions/faq/en/
WHO. (2014). Health for the world’s adolescents. A second chance in the second decade.
Summary. Geneva: Department of Maternal, Newborn, Child and Adolescent
Health.
Wiggins, S. (2004). Good for “you”: Generic and individual healthy eating advice in
family mealtimes. Journal of Health Psychology, 9(4), 35-48.
Wing, RR, Tate, DF, Gorin, AA, Raynor, HA, & Fava, JL. (2006). A self-regulation
program for maintenance of weight loss. N Engl J Med, 355, 1563 - 1571.
Witkowski, Terrence H. (2007). Food marketing and obesity in developing countries:
Analysis, ethics and public policy. Journal of Macromarketing, 27(2), 126-137.
doi: 10.1177/0276146707300076
Worsley, Anthony. (2002). Nutrition knowledge and food consumption: Can nutrition
knowledge change food behaviour? Asia Pacific J Clin Nutr, 11, 579-585.
24
2. STATE-OF-THE-ART AND THEORETICAL FRAMEWORK
This chapter reviews relevant research and provides an overview of various
approaches to consumer behaviour and behaviour change used in this thesis to understand
how adolescents’ healthy eating is learned, influenced and possibly changed.
Understanding eating behaviour is complex, and therefore the literature review
incorporates approaches and research from various fields such as health psychology,
sociology, social psychology, social marketing and communication. With a sociocognitive approach, the literature review will focus on consumer socialisation, social
norm influence and behaviour change interventions using feedback.
2.1. Consumer behaviour and behaviour change
Eating is for most of us not a choice – it is something humans do in order to
survive and therefore a vital consumption behaviour. However, with today’s extensive
supply of energy-dense food, it has become increasingly important to choose what to eat
and what not to eat.
In consumer behaviour research, food choice models have been used to
explain decision processes and the factors influencing food choice such as attitudes,
perception of sensory attributes, hunger, nutritional content (see for instance Shepherd,
1999; Shepherd & Raats, 2006) – also in relation to adolescents (Contento, Williams,
Michela, & Franklin, 2006; Shepherd & Dennison, 1996). Food choice models focus not
only on physiological or nutritional needs, but also on a range of complex and often
interrelated factors such as social, economic and cultural factors (Armitage, 2002). Even
so, food choice models do not provide a detailed account of the social factors such as
25
social norms and family dynamics to explain the origin and development of eating
behaviours which makes them less suited for this thesis.
As mentioned in the introduction, children and adolescents are highly
influenced by their parents, families and home-settings when it comes to healthy eating.
This makes consumer socialisation research relevant for studying the establishment of
and social influences on healthy eating behaviours. According to general socialisation
theory, social learning theory and cognitive development theory, consumer socialisation
is the process where children and adolescents acquire skills, knowledge, attitudes, values
and form habits and practices related to eating. Research on consumer socialisation deals
primarily with interaction within families (Ekström, 2006; Grønhøj, 2002) such as family
decision-making (Moschis & Moore, 1979; Nørgaard, 2009), family communication
(Caruana & Vassallo, 2003; Moschis, 1985), family demography (Neeley, 2005) and
family members’ influence (Foxman, Tansuhaj, & Ekström, 1989), but also with friends
(Lau, Quadrel, & Hartman, 1990), school (Kubik, Lytle, Hannan, Perry, & Story, 2003)
and other external entities such as media, advertisements and governments (Chan,
Prendergast, Grønhøj, & Bech-Larsen, 2010).
Consumer socialisation research has been criticized for having a too
simplistic framework for studying the complexity of consumer socialisation (Aldous &
McLeod, 1974) and too little variation in the methods used (Kuhlmann, 1983). Since
then, frameworks based on stages in cognitive development have been suggested (e.g.
John, 1999) in order to explain how children relate social interaction and the co-existence
of several socialisation agents to consumer behaviour. However, it seems that the multiple
roots of consumer socialisation in itself is a challenge, since “it is difficult to distinguish
consumer socialisation from other processes of change and learning” (Ekström, 2006, p.
72).
26
Besides the shaping of and influences on healthy eating behaviour, this
thesis also aims at understanding behaviour changes in adolescent healthy eating.
Consumer socialisation research explains how individuals acquire knowledge, skills etc.
to perform healthy eating behaviours. However, more details are needed to understand
predictors and principles on how to inform, enable, guide and motivate adolescents in
behaviour change. Hence, social psychology approaches to healthy eating are useful.
Models such as Theory of Reasoned Action (TRA) (Fishbein & Ajzen, 1975), Theory of
Planned Behaviour (TPB) (Ajzen, 1991) and Bandura’s Social Cognitive Theory (SCT)
(Bandura, 1986) not only look at the actual behaviour, but also the perceived antecedents
of behaviour where intention to perform a behaviour is seen as the closest cognitive
antecedent of actual behaviour. The social cognitive perspective is regarded the most
prevalent approach in the literature to study eating as a behaviour (Armitage, 2002), and
both Theory of Planned Behaviour and Social Cognitive Theory have been applied
extensively in studying various health-related behaviours (Conner & Norman, 2005).
Both models have served as a theoretical foundation for behaviour change interventions
targeting healthy eating (see for instance Cerin, Barnett, & Baranowski, 2009; Conner,
Norman, & Bell, 2002; Grønhøj, Bech-Larsen, Chan, & Tsang, 2013; Hardeman et al.,
2002; Povey, Conner, Sparks, James, & Shepherd, 2000). By comparing the main sociocognitive determinants and their areas of overlap in different conceptual models of health
behaviour such as TRA, TPB, Protection Motivation Theory, Health Belief Model and
SCT, Bandura shows that SCT is the most comprehensive model (Bandura, 2004, p. 147).
However, as the inventor of SCT, he is not unbiased.
The SCT approach to adolescent healthy eating is adopted in this thesis as
the integrating theoretical framework, because it incorporates the motivational aspects of
behaviour change and the social learning aspects which can explain the shaping of and
27
social influences on behaviour. The next two sections will outline the social learning
aspects and the behaviour change aspects of SCT in relation to adolescent healthy eating.
2.2. A socio-cognitive theory approach to consumer socialisation
In the field of marketing, consumer socialisation is defined as “processes by
which young people acquire skills, knowledge and attitudes relevant to their functioning
as consumers in the marketplace” (Ward, 1974, p. 2). This definition originates from the
general understanding of socialisation which across different social science disciplines
refers to the lifelong process of acquiring, inheriting and disseminating skills, norms,
customs and values – all needed in order to participate in and be a member of society
(Berger & Luckmann, 1966). Socialisation researchers seem to agree that internalisation
of values without external sanctions is the ultimate goal of most socialisation efforts
(Berger & Luckmann, 1966; Ekström, 2006; Maccoby, 1992). However, this is not
always the case and compliance is frequently the short-term goal of socialisation efforts
(Ekström, 2006). Consumer socialisation in terms of healthy eating is about acquiring
knowledge, skills and attitudes about healthy and unhealthy foods and adapting healthy
eating practices and habits.
According to Moschis and Churchill, consumer socialisation research is
primarily based on two models of human learning: cognitive development models and
social learning models (1978). Cognitive development models based on Piaget’s theory of
cognitive development and information processing theory relate children’s cognitive
abilities to consumer socialisation. According to John’s conceptual framework (see p. 186
for an overview, John, 1999), children aged 3-7 years are in the perceptual stage with a
general orientation toward the immediate and readily observable. Children aged 7-11 are
at the analytical stage, where information processing abilities are improved, more
28
complex knowledge is taken in and new perspectives that go beyond their own feelings
and motives can be applied. At the reflective stage (age 11-16), the social and cognitive
abilities are further developed, and “attempts to influence parents and friends reflect more
social awareness as adolescents become more strategic, favouring strategies that they
think will be better received than a simple direct approach” (John, 1999, p. 187). This
stage-model approach is operational, but as Moschis and Churchill underline, age as an
index of cognitive development can only be used in a predictive, not in an explanatory,
sense (1978).
Socialisation has been divided into primary socialisation, which takes
places during childhood, and secondary socialisation, which is the subsequent
socialisation (Berger & Luckmann, 1966), thereby corresponding to the cognitive
development stages. Consumer socialisation research has placed parents as the main
agents in the primary socialisation process, where they explicitly and implicitly transmit
consumer-related orientations to the child (Ekström, 2006; John, 1999; Ward, 1974). In
the case of healthy eating, the importance of parents in children’s healthy eating is
broadly recognised (Birch & Fisher, 1998; Kremers, Brug, de Vries, & Engels, 2003; Lau
et al., 1990) and parental influence in childhood seems to have a long-term effect (K. W.
Bauer, Laska, Fulkerson, & Neumark-Sztainer, 2011; Lake et al., 2004). In the secondary
socialisation process, agents such as peers, school and media become more influential
than in the primary socialisation process (Dotson & Hyatt, 2005; John, 1999; Moschis,
1985).
Cognitive development models may explain the “what” (skills, attitudes,
behaviour) and “who” (socialisation agents), but not necessarily the “how” and “why” of
consumer socialisation. Social modelling and social norm influence can provide insight
into how the consumer socialisation regarding healthy eating takes place. Social norms
29
have been defined as “rules and standards that are understood by members of a group and
that guide and/or constrain social behaviour without the force of laws” (Cialdini & Trost,
1998, p. 152). Cialdini and colleagues distinguish between descriptive and injunctive
norms (1991; 1990). Descriptive norms refer to what is commonly done, while injunctive
norms refer to commonly held perceptions of do’s and don’ts. In the context of SCT, it is
not so much other people’s objective behaviour or expectations as the individual’s
subjective perception of these realities that is assumed to influence behaviour (Thøgersen,
2008). Socialisation agents influence adolescent healthy eating through social norms and
modelling. Such social influences are important determinants of eating behaviour, where
people use others’ eating as a guide for what and how much to eat. Depending on how
much the others eat and the extent to which one is eager to impress them, social norms
can account for either increased or decreased intake in the presence of others (Herman,
Roth, & Polivy, 2003). It has been found that social norms influence food choices and the
quantity of food eaten (Robinson, Thomas, Aveyard, & Higgs, 2014), and that they
provide people with a sense of safety, affiliation and approval (Higgs, 2015). A review by
Cruwys et al. (2015) found that 64 out of 69 reviewed studies on social modelling and
eating documented statistically significant modelling effects suggesting that modelling is
motivated by both affiliation and uncertainty-reduction goals. Studies of social influences
on adolescent healthy eating often focus on either the influence of parents (e.g. De
Bourdeaudhuij, 1997; Neeley, 2005; Patrick & Nicklas, 2005) or of friends/peers (e.g.
Bruening et al., 2012; Howland, Hunger, & Mann, 2012; Salvy, de la Haye, Bowker, &
Hermans, 2012).
While focusing on the social influence from primary and secondary
socialisation agents, the consumer socialisation literature does not pay much attention to
how social norms from these possibly “competing” social spheres influence adolescents
30
healthy eating. A few studies compare the influence from parents and friends, such as
Pelletier et. al’s study (2014), according to which dietary behaviour appears to reflect
perceptions of normative behaviour, particularly among friends. However, the target
group is not adolescents, but college students. Salvy et. al (2011) compare the effects of
mothers and friends on children’s and adolescents’ energy intake from healthy and
unhealthy snacks. Children (age 5 to 7) are found to consume similar amounts of healthy
snacks with their friends as with their mothers, while adolescent girls consume more
healthy foods and less unhealthy food in the presence of their friends than in the company
of their mothers. This might indicate that affiliation with friends is more important than
the affiliation with parents in adolescence. However, this comparison of mothers’ and
friends’ influence did not control for the adolescents’ personal motivation for healthy
eating. Chapter 6 provides a comparison of the adolescent’s personal motivation and the
perceived descriptive and injunctive norms from both parents and friends thereby adding
to the literature by shedding light on the entanglement of social spheres in adolescent
healthy eating socialisation.
To sum up, cognitive development models help explain the “what” and
“who” of consumer socialisation, while literature on social influences can explain “how”
it takes place. It is also important to understand “why” consumer socialisation takes place.
According to John (1999), there are significant gaps in the conceptualization and
understanding of the role that social environment and experiences play in consumer
socialisation. In her review of consumer socialisation research, Ekström (2006) appeals –
based on the, in her view, too strong focus on cognitive learning in consumer socialisation
research – to a broadening of scope and methods (interpretive rather than positivistic) in
order to understand how the consumer relates to shifting consumption norms and
ideologies as well as technological changes in society.
31
Based on this and other research, it is argued in this thesis that cognitive
development models have focused too narrowly on explaining the outcomes (levels of
knowledge, skills) and the roots of social influence (socialisation agents), while social
modelling/social learning models have focused on the mechanisms of social influence
without connecting these findings in an adequate way.
Social Cognitive Theory provides a framework for integrating the cognitive
development and the social modelling aspects of consumer socialisation to a larger extent.
SCT has an agentic approach to self-development, adaptation and change (Bandura,
2001) which can help explain the “why” of consumer socialisation. According to Bandura
“people create social systems to organize, guide, and regulate human activities. The
practices of social systems, in turn, impose constraints and provide resources and
opportunity structures for personal development and functioning” (2005, p. 10). This also
goes for children and adolescents when taking their cognitive abilities into account.
SCT states that human functioning is rooted in social systems and rejects a
dualism between personal agency and a social structure disembodied from human
activity. Despite the fact that the human agency aspect “has centered almost exclusively
on personal agency exercised individually” (Bandura, 2005, p. 26), it is important to
understand that human agency operates within an interdependent causal structure
involving triadic reciprocal causation between personal factors (cognitive, affective and
biological events), environment (physically external factors such as family context and
food culture) and behaviour (see figure 1).
32
Figure 1
The “triadic model”, adapted from Bandura (1986).
The triadic model underlines that human agency cannot be fully understood
solely in terms of either social structural factors or psychological factors (which is in line
with Giddens “duality of structure” (Giddens, 1984)). Thus, socio-structural and personal
determinants are treated as interacting cofactors within a unified causal structure. The
factors do not possess equal strength, since their relative influence vary for different
activities and under different circumstances (Bandura, 1997). An example of this is the
observation that parental influence declines or at least changes as the child moves into
adolescence (Gitelson & McDermott, 2006), and the influence of friends and media
becomes stronger.
Bandura studied – inspired by Miller and Dollard (1941) – observational
learning and social modelling (Bandura, 1977b). He rejected the behaviouristic approach
to learning as a trial-and-error process with rewards and punishments (Bandura, 2005)
33
and found that social learning is not mere imitation, since observers rarely pattern their
behaviour exclusively after a single source, but combine various features of different
behaviour models to generate new versions of the behaviour or tailor it to changing
circumstances. This challenges the rather authoritarian (or naïve) view on consumer
socialisation as a uni-directional process, where children and adolescents learn and adapt
consumption related matters directly from a socialisation agent. Today’s more open and
democratic family interaction involving discussions and negotiations implies that children
and adolescents play a bigger role in household decision-making (Caruana & Vassallo,
2003; Dotson & Hyatt, 2005; Easterling, Miller, & Weinberger, 1995; Foxman et al.,
1989; Grønhøj, 2002; Nørgaard, 2009; Rose, Boush, & Shoham, 2002). Therefore,
consumer socialisation researchers have suggested that consumer socialisation is not
merely a one-way process, but can be “reverse” (Foxman et al., 1989; Grønhøj, 2002;
Moschis, 1985; Ward, 1974) or a collection of “bidirectional interactive processes”
(Kuczynski & Parkin, 2006), where mutual influence and value exchange take place
between parents and children (De Mol & Buysse, 2008; Knafo & Galansky, 2008).
However, SCT adds to this understanding that these processes happen because of the
children’s agency based on their cognitive abilities. This approach is applied in Chapter 5,
where family members’ roles and interaction regarding healthy eating socialisation is
assessed. The understanding of people as actors as well as products of their environment
is applied throughout this thesis.
2.3. A socio-cognitive theory approach to behaviour change
As explained in the previous section, human agency is central in SCT. This
section will go into detail about the elements of human agency and link them to behaviour
change.
34
Intentionality, forethought and self-regulation are core features of human
agency, since people form intentions that include action plans and strategies for realizing
them. They set themselves goals and anticipate likely outcomes. Based on studies of
guided mastery (in order for people to overcome fears such as arachnophobia), Bandura
found that competences, coping skills and self-beliefs were cultivated and enabled people
to exercise control through a feeling of self-efficacy (Bandura, 2005). Self-efficacy is a
key construct in SCT and refers to “belief in one’s capabilities to organize and execute
the courses of action required to produce given attainments” (Bandura, 1997, p. 3). Four
sources of enhancing self-efficacy have been identified with personal accomplishment or
mastery as the strongest. The other sources are vicarious experience (if a “role-model”
successfully masters a difficult situation, social comparison processes can enhance selfefficacy beliefs), verbal persuasion by others and emotional arousal (Luszczynska &
Schwarzer, 2005). In terms of behaviour change, self-efficacy can influence through the
emotions that might right arise while pursuing a goal. Self-efficacy makes behavioural
change possible since if people believe they can take action to solve a problem
instrumentally, they become more inclined to do so and feel more committed to the
decision (Luszczynska & Schwarzer, 2005). However, self-efficacy is based on the
experience “that one has the power to produce desired changes by one’s actions”
(Bandura, 2004, p. 144) and is not the same as unrealistic optimism.
While self-efficacy refers to personal action control or agency, another key
construct in SCT is outcome expectations, which refers to the anticipation of possible
consequences of one’s actions. They are organized along three dimensions: a) area of
consequences, b) positive vs. negative consequences which serve as incentives and
disincentives, and c) short-term vs. long-term consequences. They can take the form of
physical outcome expectations (e.g. weight loss when eating healthier), social (e.g.
35
receiving approval from parents when eating healthier) and self-evaluative expectations
(e.g. being ashamed for not being able to change behaviour) regarding the outcome of an
anticipated behaviour (Bandura, 1977a, 1997).
SCT states that self-efficacy beliefs and outcome expectations are
conditionally related as direct predictors of behaviour and affect goal settings and the
perception of socio-structural factors. The functionalist view that behaviour is regulated
solely by external rewards and punishments is thereby dismissed, since people display
considerable self-direction in the face of competing influences (Bandura, 1997). Figure 2
presents a model for how these elements of SCT are linked and how they influence
behaviour and behaviour change.
Figure 2
An illustration of social cognitive theory (2004)
Outcome expectations



Physical
Social
Self-evaluative
Self-efficacy
Goals (Intentions)
Socio-structural
factors


Facilitators
Barriers
36
Behaviour
Both self-efficacy beliefs and outcome expectancies play influential roles in
behaviour change and are linked to goal-setting (Luszczynska & Schwarzer, 2005). If you
have a high level of self-efficacy and positive outcome expectancies for eating healthy, it
would be easier for you to achieve your goal, even if you foresee barriers or experience
negative social influence. The importance of self-efficacy and outcome expectations for
adolescents’ healthy eating has been confirmed empirically (see for instance Fitzgerald,
Heary, Kelly, Nixon, & Shevlin, 2013) and specifically the importance of self-efficacy for
increasing adolescents’ fruit and vegetable intake has been documented by several studies
(Bere & Klepp, 2004; Geller & Dzewaltowski, 2010; Young, Fors, Fasha, & Hayes,
2004). The theoretical framework of SCT is applied in Chapter 6 in order to understand
the impact of adolescents’ self-efficacy and outcome expectations together with perceived
social norms from parents and friends (environment).
The next section goes into detail with the important elements of behaviour
change in the context of an intervention using feedback on the platform of mobile phones’
short message system (SMS) for promoting adolescent healthy eating.
2.4. Feedback interventions
When conducting healthy eating interventions, it is recommended to build
on a theoretical framework in order to be able to evaluate the mechanisms of change and
their usefulness in developing more effective interventions. However, even with a
theoretical framework, there is little information about how to “translate” theory into
practice and develop theory-based interventions (Michie, Johnston, Francis, Hardeman, &
Eccles, 2008). SCT is considered an exception, since it specifies how to change the main
causal determinants of behaviour, namely self-efficacy and outcome expectations (Michie
37
et al., 2008). A review by Contento et al. (2002) confirms this by showing that changing
SCT constructs (i.e. self-efficacy and outcome expectations) compared to preferences or
attitudes are more likely to produce changes in behaviour.
Luszczynska and Schwarzer (2005) describe behaviour change as a
competent self-regulation process in which individuals monitor their responses to
demanding situations, observe similar others facing similar demands, appraise their
coping resources, create optimistic self-beliefs, plan a course of action, perform the
critical action, and evaluate the outcomes. Goal-setting and human self-motivation are
important for behaviour change which relies on – according to Bandura – “discrepancy
production as well as on discrepancy reduction” (1989, p. 1180). People motivate
themselves by creating discrepancy when setting a challenging goal. Then they mobilize
their efforts on an estimation of what it would take to accomplish it. When striving to
achieve the desired results, feedback control comes into play in the subsequent
adjustments of effort (Bandura, 1989). Kluger and DeNisi define feedback interventions
as “actions taken by (an) external agent(s) to provide information regarding some
aspect(s) of one’s task performance” (1996, p. 255). Personal and timely feedback on
behaviour via information or exercises is an effective way to make participants reflect on
their own behaviour (Schultz, 1999) and enhance their self-efficacy in such a process
(Kluger & DeNisi, 1996; Luszczynska, Tryburcy, & Schwarzer, 2007). Many selfregulation theories are founded on negative feedback control systems (e.g. Carver &
Scheier, 1982), where a reduction in the perceived discrepancy between an inner
reference standard and performance feedback automatically triggers adjustments to
reduce the incongruity. However, according to Bandura, “goal adjustments do not follow
a neat pattern of ever-rising standards after personal accomplishments, nor do failures
necessarily lower aspirations” (1997, p. 132). Hence, feedback is not a guarantee for
38
positive results (Grønhøj & Thøgersen, 2011). The interacting cognitive and affective
factors can make some become less sure of their efficacy or lose faith in their capabilities,
when failing to eat healthier, while others have more “resistant” self-efficacy beliefs,
where failures may ignite more self-regulation in order to accomplish the goal. Hence,
perceived self-efficacy plays a role when choosing “what challenges to undertake, how
much effort to expend in the endeavour, and how long to persevere in the face of
difficulties” (Bandura, 1989, p. 1180), and it is acknowledged that feedback may
strengthen perceived self-efficacy with respect to performing behaviour successfully
(Bandura, 1986).
Interventions using feedback to participants can be useful in improving
performance (Kluger & DeNisi, 1996) and this mechanism have gained ground in health
behaviour interventions in recent years (see for instance De Bourdeaudhuij & Brug, 2000;
Oenema & Brug, 2003). When planning this PhD project and conducting the empirical
studies, mobile phones’ Short Messaging System (SMS, text messaging) was one of
adolescents’ preferred communication tools (Davie, Panting, & Charlton, 2004; Faulkner
& Culwin, 2005; Phau & Teah, 2009). Therefore, it seems paradoxical that only a few
studies on community-based health interventions targeting adolescents have used text
messaging. The ease and accessibility of text messaging, being affordable and available to
basically all social groups (Fjeldsoe, Marshall, & Miller, 2009), along with the ability to
receive and send brief, private, and personalized messages or reminders (Preston,
Walhart, & O'Sullivan, 2011), makes text messaging an obvious choice for an interactive
feedback intervention. The rapid development in communication technology with
smartphone applications providing immediate tailored feedback in a more aesthetic and
detailed manner than text messages (Bort-Roig, Gilson, Puig-Ribera, Contreras, & Trost,
2014) might overhaul text messaging as a health promoting tool – future studies will tell.
39
However, for the purpose of this thesis, the technological platform for health promotion is
not as important as is the theoretical underpinnings of behaviour change and the role of
immediate feedback.
Employing tailored feedback via text messaging has been used in clinical
settings with adults (see for instance de Jongh, Gurol-Urganci, Vodopivec-Jamsek, Car,
& Atun, 2012; Krishna, Boren, & Balas, 2009; Nobis et al., 2013; Nundy, Dick, Solomon,
& Peek, 2013). Feedback interventions using text messaging and targeting adolescents’
health behaviours have focused on cancer prevention (Lana, del Valle, Lopez, FayaOrnia, & Lopez, 2013), smoking cessation (Haug, Meyer, Schorr, Bauer, & John, 2009),
intake of sugar-sweetened beverages, physical activity and screen time (Shapiro et al.,
2008), consumption of breakfast, fruit and vegetables, sweetened beverages, fast food and
decreased screen time (Woolford, Clark, Strecher, & Resnicow, 2010), eating and
exercise habits (S. Bauer, de Niet, Timman, & Kordy, 2010) and fruit and vegetable
consumption (Bech-Larsen & Grønhøj, 2013). The results are not impressive – for
instance did Bech-Larsen and Grønhøj (2013) find that adolescents (age 10-12) with low
pre-intervention intake of fruit and vegetables had a small, but significant increase in
consumption, whereas children with high pre-intervention intake (unexpectedly) reduced
their consumption during the intervention.
Chapter 7 provides new insights on the use of text messaging in a feedback
intervention targeting adolescents’ fruit and vegetable intake. Using the SCT framework,
the importance of feedback for behaviour change and changes in the antecedents of
behaviour, namely self-efficacy and outcome expectations, are discussed.
40
2.5. Theoretical approach – overview
With the aim of understanding adolescent health eating and possible
behaviour change, it is crucial to have a theoretical framework that can help explain both.
The choice of socio-cognitive theory as a theoretical framework allows the study of how
behaviours are learned and influenced, but also the study of behaviour change using
feedback.
When studying consumer behaviours such as healthy eating, it is important
to understand how these behaviours come about and are influenced. By adapting a sociocognitive approach to consumer socialisation, this chapter has demonstrated that
consumer socialisation does not have to be a choice between cognitive development
theory and social learning theory, since socio-cognitive theory can incorporate both
aspects and thereby provide a deeper and more integrated insight into adolescent healthy
eating. Hence, using this theoretical framework to understand adolescent healthy eating, a
more integrated view on behaviour establishment and behaviour change is provided.
41
References
Ajzen, Icek. (1991). The theory of planned behavior. Organizational Behavior and
Human Decision Processes, 50, 179-211.
Aldous, Joan, & McLeod, Jack M. (1974). Commentaries on Ward, "Consumer
Socialization". Journal of Consumer Research, 1(2), 15-17. doi: 10.2307/2489101
Armitage, Christopher J. (2002). The social psychology of food. Buckingham: Open
University Press.
Bandura, Albert. (1977a). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84, 191-215.
Bandura, Albert. (1977b). Social learning theory (1st ed.). Englewood Cliffs, NJ:
Prentice-Hall.
Bandura, Albert. (1986). Social foundations of thought and action: A social cognitive
perspective. Englewood Cliffs, NJ: Prentice-Hall.
Bandura, Albert. (1989). Human agency in social cognitive theory. American
Psychologist, 44(9), 1175-1184. doi: 10.1037/0003-066x.44.9.1175
Bandura, Albert. (1997). Self-efficacy: The exercise of control. New York: W. H.
Freeman.
Bandura, Albert. (2001). Social cognitive theory: An agentic perspective. Annual Review
of Psychology, 52, 1-26.
Bandura, Albert. (2004). Health promotion by social cognitive means. Health Education
& Behavior 31(2), 143-164.
Bandura, Albert. (2005). The evolution of social cognitive theory. In K. G. H. Smith, M.
A. (Ed.), Great minds in management (pp. 9-35). Oxford: Oxford University
Press.
Bauer, K W, Laska, Melissa N, Fulkerson, Jayne A, & Neumark-Sztainer, D. (2011).
Longitudinal and secular trends in parental encouragement for healthy eating,
physical activity and dieting throughout the adolescent years. Journal of
Adolescent Health, 49(3), 306. doi: 10.1016/j.jadohealth.2010.12.023
Bauer, S, de Niet, J, Timman, R, & Kordy, H. (2010). Enhancement of care through selfmonitoring and tailored feedback via text messaging and their use in the treatment
of childhood overweight. Patient Education and Counseling, 79, 313-319.
42
Bech-Larsen, Tino, & Grønhøj, Alice. (2013). Promoting healthy eating to children: A
text message (SMS) feedback approach. International Journal of Consumer
Studies, 37(3), 250-256. doi: 10.1111/j.1470-6431.2012.01133.x
Bere, Elling, & Klepp, Knut-Inge. (2004). Correlates of fruit and vegetable intake among
Norwegian schoolchildren: Parental and self-reports. Public Health Nutrition,
7(8), 991-998. doi: 10.1079/phn2004619
Berger, Peter L., & Luckmann, Thomas. (1966). The social construction of reality: A
treatise in the sociology of knowledge. Garden City, NY: Doubleday.
Birch, L., & Fisher, J. (1998). Development of eating behaviors among children and
adolescents. Pediatrics, 101(3), 539-549.
Bort-Roig, Judit, Gilson, Nicholas D, Puig-Ribera, Anna, Contreras, Ruth S, & Trost,
Stewart G. (2014). Measuring and influencing physical activity with smartphone
technology: A systematic review. Sports Medicine, 44(5), 671-686. doi:
10.1007/s40279-014-0142-5
Bruening, M. , Eisenberg, M. , MacLehose, R. , Nanney, M. , Story, M. , & NeumarkSztainer, D. (2012). Relationship between adolescents’ and their friends’ eating
behaviors: Breakfast, fruit, vegetable, whole-grain and dairy intake. Journal of
Academy of Nutrition and Dietetics, 112, 1608-1613. doi:
10.1016/j.jand.2012.07.008
Caruana, Albert, & Vassallo, Rosella. (2003). Children's perception of their influence
over purchases: The role of parental communication patterns. Journal of
Consumer Marketing, 20(1), 55-66.
Carver, Charles S., & Scheier, Michael F. (1982). Control theory: A useful conceptual
framework for personality-social, clinical and health psychology. Psychological
Bulletin, 92(1), 111-135.
Cerin, Ester, Barnett, Anthony, & Baranowski, Tom. (2009). Testing theories of dietary
behavior change in youth using the mediating variable model with intervention
programs. Journal of Nutrition Education and Behavior, 41(5), 309-318. doi:
10.1016/j.jneb.2009.03.129
Chan, Kara, Prendergast, Gerard, Grønhøj, Alice, & Bech-Larsen, Tino. (2010). The role
of socializing agents in communicating healthy eating to adolescents: A cross
cultural study. Journal of International Consumer Marketing, 23(1), 59–74. doi:
10.1080/08961530.2011.524578
Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of normative
conduct: A theoretical refinement and re-evaluation of the role of norms in human
behavior. Advances in Experimental Social Psychology, 24, 201-234. doi:
10.1016/S0065-2601(08)60330-5
43
Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative
conduct: Recycling the concept of norms to reduce littering in public places.
Journal of Personality and Social Psychology, 58, 1015-1026. doi:
10.1037//0022-3514.58.6.1015
Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity and
compliance. In D. T. Gilbert, S. T. Fiske & G. Lindzey (Eds.), The handbook of
social psychology (4 ed., Vol. 2, pp. 151–192 ). New York: McGraw-Hill.
Conner, Mark, & Norman, Paul. (2005). Predicting health behaviour: Research and
practice with social cognition models (2nd ed.). Berkshire, GBR: McGraw-Hill
Education.
Conner, Mark, Norman, Paul, & Bell, Russell. (2002). The theory of planned behavior
and healthy eating. Health Psychology, 21(2), 194-201. doi: 10.1037/02786133.21.2.194
Contento, Isobel R., Randell, Jill S., & Basch, Charles E. (2002). Review and analysis of
evaluation measures used in nutrition education intervention research. Journal of
Nutrition Education and Behavior, 34(1), 2-25. doi: 10.1016/S14994046(06)60220-0
Contento, Isobel R., Williams, Sunyna S., Michela, John L., & Franklin, Amie B. (2006).
Understanding the food choice process of adolescents in the context of family and
friends. Journal of Adolescent Health, 38(5), 575-582. doi:
10.1016/j.jadohealth.2005.05.025
Cruwys, Tegan, Bevelander, Kirsten E, & Hermans, Roel C.J. (2015). Social modeling of
eating: A review of when and why social influence affects food intake and choice.
Appetite, 86, 3. doi: 10.1016/j.appet.2014.08.035
Davie, Ronald, Panting, Charlotte, & Charlton, Tony. (2004). Mobile phone ownership
and usage among pre-adolescents. Telematics and Informatics, 21(4), 359-373.
doi: 10.1016/j.tele.2004.04.001
De Bourdeaudhuij, Ilse. (1997). Perceived family members' influence on introducing
healthy food into the family. Health Education Research, 12(1), 77-90.
De Bourdeaudhuij, Ilse, & Brug, Johannes. (2000). Tailoring dietary feedback to reduce
fat intake: An intervention at the family level. Health Education Research, 15(4),
449-462. doi: 10.1093/her/15.4.449
de Jongh, Thyra, Gurol-Urganci, Ipek, Vodopivec-Jamsek, Vlasta, Car, Josip, & Atun,
Rifat. (2012). Mobile phone messaging for facilitating self-management of longterm illnesses. The Cochrane database of systematic reviews, 12, CD007459.
44
De Mol, Jan, & Buysse, Ann. (2008). The phenomenology of children's influence on
parents. Journal of Family Therapy, 30(2), 163-193. doi: 10.1111/j.14676427.2008.00424.x
Dotson, Michael J., & Hyatt, Eva M. (2005). Major influence factors in children's
consumer socialization. Journal of Consumer Marketing, 22(1), 35-42.
Easterling, Debbie, Miller, Shirley, & Weinberger, Nanci. (1995). Environmental
consumerism: A process of children's socialization and families' resocialization.
Psychology and Marketing, 12, 531-550.
Ekström, Karin M. . (2006). Consumer socialization revisited. In R. Belk (Ed.), Research
in Consumer Behavior (pp. 71-98): Emerald Group Publishing Limited.
Faulkner, Xristine, & Culwin, Fintan. (2005). When fingers do the talking: A study of
text messaging. Interacting with Computers, 17(2), 167-185. doi:
10.1016/j.intcom.2004.11.002
Fishbein, Martin A., & Ajzen, Icek. (1975). Belief, attitude, intention and behavior: An
introduction to theory and research. Reading, MA: Addison-Wesley.
Fitzgerald, A., Heary, C., Kelly, C., Nixon, E., & Shevlin, M. (2013). Self-efficacy for
healthy eating and peer support for unhealthy eating are associated with
adolescents’ food intake patterns. Appetite, 63, 48-58. doi:
10.1016/j.appet.2012.12.011
Fjeldsoe, Brianna S., Marshall, Alison L., & Miller, Yvette D. (2009). Behavior change
interventions delivered by mobile telephone short-message service. American
Journal of Preventive Medicine, 36, 165-173.
Foxman, Ellen R., Tansuhaj, Patriya S., & Ekström, Karin M. (1989). Family members'
perceptions of adolescents' influence in family decision making. Journal of
Consumer Research, 15(4), 482-491.
Geller, K., & Dzewaltowski, D. (2010). Examining elementary school-aged children's
self-efficacy and proxy efficacy for fruit and vegetable consumption. Health
Education & Behavior, 37(4), 465-478. doi: 10.1177/1090198109347067
Giddens, Anthony. (1984). The constitution of society: Outline of the theory of
structuration. Cambridge: Polity Press.
Gitelson, I. , & McDermott, D. (2006). Parents and their young adult children: Transitions
to adulthood. Child Welfare, 86(5), 853-866.
Grønhøj, Alice. (2002). Miljøvenlig adfærd i familien: Et studie af familiemedlemmernes
involvering og sociale interaktion [Pro-environmental behaviour in the family: A
45
study of family members' involvement and social interaction]. Århus: Institut for
Markedsøkonomi.
Grønhøj, Alice, Bech-Larsen, Tino, Chan, Kara, & Tsang, Lennon. (2013). Using theory
of planned behavior to predict healthy eating among Danish adolescents. Health
Education, 113(1), 4-17. doi: 10.1108/09654281311293600
Grønhøj, Alice, & Thøgersen, John. (2011). Feedback on household electricity
consumption: Learning and social influence processes. International Journal of
Consumer Studies, 35(2), 138-145. doi: 10.1111/j.1470-6431.2010.00967.x
Hardeman, Wendy, Kinmonth, Ann Louise, Johnston, Marie, Wareham, Nicholas,
Johnston, Derek, & Bonetti, Debbie. (2002). Application of the theory of planned
behaviour in behaviour change interventions: A systematic review. Psychology &
Health, 17(2), 123-158. doi: 10.1080/08870440290013644a
Haug, Severin, Meyer, Christian, Schorr, Gudrun, Bauer, S, & John, Ulrich. (2009).
Continuous individual support of smoking cessation using text messaging: A pilot
experimental study. Nicotine & tobacco research: Official Journal of the Society
for Research on Nicotine and Tobacco, 11(8), 915-923.
Herman, C. Peter, Roth, Deborah A, & Polivy, Janet. (2003). Effects of the presence of
others on food intake: A normative interpretation. Psychological bulletin, 129(6),
873-886. doi: 10.1037/0033-2909.129.6.873
Higgs, Suzanne. (2015). Social norms and their influence on eating behaviours. Appetite,
86(0), 38-44. doi: 10.1016/j.appet.2014.10.021
Howland, M. , Hunger, J. , & Mann, T. (2012). Friends don’t let friends eat cookies:
Effects of restrictive eating norms on consumption among friends. Appetite, 59,
505-509. doi: 10.1016/j.appet.2012.06.020
John, Deborah Roedder. (1999). Consumer socialization of children: A retrospective look
at twenty-five years of research. Journal of Consumer Research, 26(3), 183-213.
Kluger, A.N., & DeNisi, A. (1996). The effects of feedback interventions on
performance: A historical review, a meta-analysis and a preliminary feedback
intervention theory. Psychological Bulletin, 119(2), 254-284.
Knafo, Ariel, & Galansky, Neta. (2008). The influence of children on their parents’
values. Social and Personality Psychology Compass, 2(3), 1143-1161. doi:
10.1111/j.1751-9004.2008.00097.x
Kremers, Stef P. J., Brug, Johannes, de Vries, Hein, & Engels, Rutger C. M. E. (2003).
Parenting style and adolescent fruit consumption. Appetite, 41(1), 43-50.
46
Krishna, Santosh, Boren, Suzanne Austin, & Balas, E Andrew. (2009). Healthcare via cell
phones: A systematic review. Telemedicine journal and e-health: The official
journal of the American Telemedicine Association, 15(3), 231-240. doi:
10.1089/tmj.2008.0099
Kubik, Martha Y., Lytle, Leslie A., Hannan, Peter J., Perry, Cheryl L., & Story, Mary.
(2003). The association of the school food environment with dietary behaviors of
young adolescents. Am J Public Health, 93(7), 1168-1173. doi:
10.2105/ajph.93.7.1168
Kuczynski, L., & Parkin, M. C. (2006). Agency and bidirectionality in socialization. In J.
E. Grusec & P. D. Hastings (Eds.), Handbook of socialization: Theory and
research (pp. 259-283). New York: The Guilford Press.
Kuhlmann, Eberhard. (1983). Consumer socialization of children and adolescents: A
review of current approaches. Journal of Consumer Policy, 6, 397-418.
Lake, A. A., Rugg-Gunn, Andrew J., Hyland, Rob M., Wood, Charlotte E., Mathers, John
C., & Adamson, Ashley J. (2004). Longitudinal dietary change from adolescence
to adulthood: Perceptions, attributions and evidence. Appetite, 42(3), 255-263.
doi: 10.1016/j.appet.2003.11.008
Lana, Alberto, del Valle, Maria Olivo, Lopez, Santiago, Faya-Ornia, Goretti, & Lopez,
Maria Luisa. (2013). Study protocol of a randomized controlled trial to improve
cancer prevention behaviors in adolescents and adults using a web-based
intervention supplemented with SMS. BMC Public Health, 13(1), 357.
Lau, Richard R., Quadrel, Marilyn Jacobs, & Hartman, Karen A. (1990). Development
and change of young adults' preventive health beliefs and behavior: Influence
from parents and peers. Journal of Health and Social Behavior, 31(3), 240-259.
Luszczynska, A., & Schwarzer, R. (2005). Social cognitive theory. In M. Conner & P.
Norman (Eds.), Predicting health behaviour: Research and practice with social
cognition models (2nd ed., pp. 127–169). Berkshire: Open University Press.
Luszczynska, A., Tryburcy, M., & Schwarzer, R. (2007). Improving fruit and vegetable
consumption: A self-efficacy intervention compared to a combined self-efficacy
and planning intervention. Health Education Research, 22, 630-638.
Maccoby, E. E. (1992). The role of parents in the socialization of children: An historical
overview. Developmental Psychology, 28(6), 1006-1017.
Michie, Susan, Johnston, Marie, Francis, Jill, Hardeman, Wendy, & Eccles, Martin.
(2008). From theory to intervention: Mapping theoretically derived behavioural
determinants to behaviour change techniques. Applied Psychology, 57(4), 660680. doi: 10.1111/j.1464-0597.2008.00341.x
47
Miller, N. E., & Dollard, J. (1941). Social learning and imitation. New Haven, CT, US: Yale
University Press.
Moschis, G. (1985). The role of family communication in consumer socialization of
children and adolescents. Journal of Consumer Research, 11(4), 898-913.
Moschis, G., & Churchill, Gilbert A. (1978). Consumer socialization: A theoretical and
empirical analysis. Journal of Marketing Research, 15, 599-609.
Moschis, G., & Moore, Roy L. (1979). Decision making among the young: A
socialization perspective. Journal of Consumer Research, 6(2), 101-112.
Neeley, Sabrina. (2005). Influences on consumer socialisation. Young Consumers, 6(2),
63-69. doi: 10.1108/17473610510701115
Nobis, Stephanie, Lehr, Dirk, Ebert, David Daniel, Berking, Matthias, Heber, Elena,
Baumeister, Harald, . . . Riper, Heleen. (2013). Efficacy and cost-effectiveness of
a web-based intervention with mobile phone support to treat depressive symptoms
in adults with diabetes mellitus type 1 and type 2: Design of a randomised
controlled trial. BMC psychiatry, 13(1), 306-306. doi: 10.1186/1471-244x-13-306
Nundy, Shantanu, Dick, Jonathan J, Solomon, Marla C, & Peek, Monica E. (2013).
Developing a behavioral model for mobile phone-based diabetes interventions.
Patient Education and Counseling, 90(1), 125.
Nørgaard, Maria Kümpel. (2009). Family decision-making during food buying. (PhD
thesis), MAPP Centre, Aarhus School of Business, Aarhus University.
Oenema, Anke, & Brug, Johannes. (2003). Feedback strategies to raise awareness of
personal dietary intake: Results of a randomized controlled trial. Preventive
Medicine, 36(4), 429-439. doi: 10.1016/S0091-7435(02)00043-9
Patrick, Heather, & Nicklas, Theresa A. (2005). A review of family and social
determinants of children’s eating patterns and diet quality. Journal of the
American College of Nutrition, 24(2), 83-92. doi:
10.1080/07315724.2005.10719448
Pelletier, Jennifer E, Graham, Dan J, & Laska, Melissa N. (2014). Social norms and
dietary behaviors among young adults. American Journal of Health Behavior,
38(1), 144-152. doi: 10.5993/ajhb.38.1.15
Phau, I., & Teah, M. . (2009). Young consumers’ motives for using SMS and perceptions
towards SMS advertising. Direct Marketing, 3(2), 97-108.
48
Povey, R., Conner, Mark, Sparks, Paul, James, Rhiannon, & Shepherd, Richard. (2000).
Application of the theory of planned behaviour to two dietary behaviours: Roles
of perceived control and self-efficacy. Journal of Health Psychology, 5, 121-139.
Preston, Karen E, Walhart, Tara A, & O'Sullivan, Ann L. (2011). Prompting healthy
behavior via text messaging in adolescents and young adults. American Journal of
Lifestyle Medicine, 5(3), 247-252. doi: 10.1177/1559827610392325
Robinson, Eric, Thomas, Jason, Aveyard, Paul, & Higgs, Suzanne. (2014). What
everyone else is eating: A systematic review and meta-analysis of the effect of
informational eating norms on eating behavior. Journal of the Academy of
Nutrition and Dietetics, 114(3), 414-429. doi: 10.1016/j.jand.2013.11.009
Rose, Gregory M., Boush, David , & Shoham, Aviv. (2002). Family communication and
children's purchasing influence: A cross-national examination. Journal of
Business Research, 55(11), 867-873.
Salvy, Sarah-Jeanne, de la Haye, K. , Bowker, J. C. , & Hermans, R. C. J. (2012).
Influence of peers and friends on children's and adolescents' eating and activity
behaviors. Physiology & Behavior, 106, 369-378. doi:
10.1016/j.physbeh.2012.03.022
Salvy, Sarah-Jeanne, Elmo, Alison, Nitecki, Lauren A, Kluczynski, Melissa A, &
Roemmich, James N. (2011). Influence of parents and friends on children's and
adolescents' food intake and food selection. The American Journal of Clinical
Nutrition, 93(1), 87-92. doi: 10.3945/ajcn.110.002097
Schultz, P. Wesley. (1999). Changing behavior with normative feedback interventions: A
field experiment of curbside recycling. Basic and Applied Social Psychology, 21,
25-36.
Shapiro, Jennifer R., Bauer, S., Hamer, Robert M., Kordy, Hans, Ward, Dianne, & Bulik,
Cynthia M. (2008). Use of text messaging for monitoring sugar-sweetened
beverages, physical activity and screen time in children: A pilot study. Journal of
Nutrition Education and Behavior, 40(6), 385-391. doi:
10.1016/j.jneb.2007.09.014
Shepherd, Richard. (1999). Social determinants of food choice. Proceedings of the
Nutrition Society, 58(4), 807-812. doi: 10.1017/s0029665199001093
Shepherd, Richard, & Dennison, Catherine M. (1996). Influences on adolescent food
choice. Proceedings of the Nutrition Society, 55(1B), 345-357. doi:
10.1079/pns19960034
Shepherd, Richard, & Raats, Monique. (2006). The psychology of food choice.
Cambridge, MA: CABI Pub.
49
Thøgersen, John. (2008). Social norms and cooperation in real-life social dilemmas.
Journal of Economic Psychology, 29(4), 458-472. doi: 10.1016/j.joep.2007.12.004
Ward, Scott. (1974). Consumer socialization. Journal of Consumer Research, 1(2), 1-17.
Woolford, Susan J, Clark, Sarah J, Strecher, Victor J, & Resnicow, Kenneth. (2010).
Tailored mobile phone text messages as an adjunct to obesity treatment for
adolescents. Journal of Telemedicine and Telecare, 16(8), 458-461. doi:
10.1258/jtt.2010.100207
Young, E., Fors, S. W., Fasha, E., & Hayes, D. (2004). Associations between perceived
parent behaviors and middle school student fruit and vegetable consumption.
Journal of Nutrition Education and Behavior (36), 2-12.
50
3. METHODOLOGY AND RESEARCH DESIGN
This chapter presents general considerations concerning healthy eating as a
research topic and adolescents and their families as research subjects. It outlines the
methodological standpoint and research strategy chosen to explore adolescents’ healthy
eating.
3.1. Studying healthy eating
A number of studies have explored adult consumers’ perceptions of healthy
eating (Falk, Sobal, Bisogni, Connors, & Devine, 2001; Lake et al., 2007; Margetts,
Martinez, Saba, Holm, & Kearney, 1997; Povey, Conner, Sparks, James, & Shepherd,
1998). They find that consumers define healthy eating as eating limited quantities of fat,
salt and sugar, eating a balanced, varied diet, and eating natural, unprocessed foods.
According to Kazbaré (2010), the perception of healthy eating also sometimes overlap
with the perception of weight control/weight loss.
A systematic review on children and healthy eating shows that “definitions
of healthy eating vary, but an emphasis on achieving the right balance of different foods
is a common component” (Thomas et al., 2003, p. 8). This is for instance the case in Croll
et al.’s (2001) study, according to which adolescents believe that healthy eating involves
moderation, balance and variety. The adolescents had a significant amount of knowledge
regarding healthy foods, but also demonstrated a knowledge-behaviour gap (as described
in section 1.1.). When studying healthy eating, one should be aware of the possible
mismatch between knowledge, attitudes, perceived healthiness and reported or observed
behaviour. Potential social desirability effects should also be considered (Klesges et al.,
2004).
51
This PhD project investigates a specific healthy eating behaviour; the
intake of fruit and vegetables which is treated in accordance with the common
understanding of healthy eating and following the logic of the official dietary guidelines.
It is acknowledged that healthy eating reaches beyond this specific behaviour.
3.2. Studying adolescents
According to Scott (2008, p. 87) “the construction of childhood that views
children as incomplete adults is coming under attack and there is a new demand for
research that focuses on children as actors in their own right”. Research with children has
long been central to developmental psychologists and education specialists, but in for
instance surveys on the general population, children have usually been regarded out of
scope (Scott, 2008). With consumer socialisation research in the 1970’s, research
involving children became more widely used and more advanced (John, 1999). The
importance of listening to children’s perceptions as a supplement to parents’ perceptions
was recognised (Gulløv & Højlund, 2003), and children are now considered valuable
sources of information. However, involving children and adolescents in research projects
requires reflections from the researcher that differ from when involving adults and creates
certain challenges which should be responded to in an appropriate way (Banister &
Booth, 2005; Nørgaard, 2009; Scott, 2008).
Children may not be able to give relevant to-the-point answers (Mayes,
2000). When involving children in research, the researcher should consider the child’s
cognitive competences (i.e. concentration and memory) and language competences such
as communication, articulation of opinions, reading and concept and sentence
understanding (Andersen, 2000; Andersen & Kjærulff, 2003; Andersen & Ottosen, 2002;
52
John, 1999; Scott, 2008). As described in section 2.2., John’s conceptual framework
understands consumer socialisation as a series of stages, where maturity in cognitive and
social terms determines the transition between stages (John, 1999). Banister and Booth
agree that children’s development “should be recognised in social terms rather than with
sole regard to age” (2005, p. 159) – but age may be a useful indicator (Christensen &
James, 2008) of children’s capability to participate in survey studies.
Involving children of different ages in a research project can require
different types of methods for each group in order to obtain the best data (Gulløv &
Højlund, 2003; Nørgaard, 2009). This PhD project involves young (11-12 years) and
older (15-16 years) adolescents (cf. the argumentation in section 1 and 2), and it is
acknowledged that they possess different cognitive and linguistic competences. The
qualitative methods used (interviews and a practical exercise) are, however, considered
appropriate to elicit meaningful data from both age groups. With the quantitative methods
(surveys and a feedback intervention), a number of pre-tests were done in the youngest
age group. The argument is that if they can understand and answer meaningfully, so can
older adolescents and parents which makes it possible to compare results across age
groups and generations.
Ethics is always important in research, but especially in studies involving
children and adolescents (Kampmann, 1998; Nørgaard, 2009). Trustworthiness,
confidence, anonymity and professional secrecy are just as important when doing
research with children and adolescents as with adults (Tufte, 2000). The notion of consent
and the different levels of ethical responsibility involved in a study (e.g. children, parents,
teachers, researchers) should also be considered (Banister & Booth, 2005). Nørgaard
presents ten ethical guidelines for doing (especially ethnographic) research involving
children (2009, p. 73) based on Kampmann’s recommendations (Kampmann, 1998). They
53
involve considerations regarding the aim of the project, costs and benefits, privacy and
confidence, recruitment, budget, information to children and parents, consent,
presentation and consequences. These guidelines and other advice on ethics regarding
studying children and adolescents have been followed. Also, since Aarhus University has
no ethical review board to consult regarding the nature of this study, the project complied
to the official guidelines for research at MAPP (the details are presented in the design of
the empirical studies in section 3.4.).
3.3. Mixed methods approach
This PhD project applies both qualitative and quantitative research methods.
The debate about the relations between qualitative and quantitative research has been
going on for several decades at different levels such as epistemology and methodology,
research designs and methods, generalization and linking of findings and assessing the
quality of research (Denzin, 2009; Flick, 2009). Mixed methods refer to the use of two or
more methods in a research project yielding both qualitative and quantitative data, while a
combination of methods which yield data of the same kind are referred to as multimethods (Tashakkori & Teddlie, 2010). This dissertation uses both a mixed method and a
multi-method approach, since it applies both qualitative and quantitative methods, but
also different kinds of quantitative and qualitative methods, respectively, in a single
study. The mixed method approach is chosen in order to obtain a more nuanced
understanding of adolescents healthy eating based on the conviction that qualitative and
quantitative approaches result in different insights and levels of knowledge.
One advantage of quantitative methods is the ability to provide more
representative studies. A large number of respondents can be involved – often without
spending more time. This is not the case in qualitative methods. With quantitative
54
methods it is possible to measure and quantify phenomena on their frequency and
distribution and thereby allow generalization of findings (Flick, 2009). Some of the
advantages of qualitative methods compared to quantitative are the ability to provide data
and methods to interpret and understand (Silverman, 2001). Another advantage is that the
“collection of data and its analysis are sensitive to the context aiming at a holistic
understanding of the issues studied” (Eriksson & Kovalainen, 2008, p. 5). One
disadvantage of qualitative research is that it is time consuming and put heavy demands
on the qualitative researchers’ competences – and even more so when doing research with
children (Nørgaard, 2009).
Finding a rationale for combining qualitative and quantitative data has been
heavily debated in the literature on mixed methods – it has even been called a “paradigm
war” (Bryman, 2006; Hall, 2012). The range of controversies in mixed method research
goes from “basic issues of the legitimacy and meaning of mixed methods to its
philosophical underpinnings and on to the pragmatics of conducting a mixed methods
study” (Creswell, 2011, p. 281). The different stances can (simplified) be summed up by
Rossman and Wilson’s (1985) three perspectives: the purist approach where the two
methods are seen as mutually exclusive because of the incompatibility of the underlying
paradigms, the situationalist approach that views them as separate but equal, and the
pragmatist approach that suggests that integration is possible. The purist approach is
represented by Bryman (1984) who focuses on the paradigm level and is critical towards
mixed method researchers who do not reflect on the epistemology (‘methodology’) versus
the technical issues (‘methods’) in a study, while Morgan (2007, p. 48), representing the
pragmatist approach, is more concerned with “redirect[ing] our attention to
methodological rather than metaphysical concerns”. The pragmatic approach supports
mixed methods research (Creswell, 2011; Creswell & Plano Clark, 2011; Johnson,
55
Onwuegbuzie, & Turner, 2007) since it rejects traditional dualisms, advocates a needsbased approach to research method and concept selection, endorses pluralism and judges
the workability of research designs and theories based on the criteria of predictability and
applicability (Johnson & Onwuegbuzie, 2004). The pragmatic approach even caused
“paradigm peace” to be declared (Bryman, 2006).
Mixing methods in a single study is basically triangulation. Denzin defines
triangulation as “the combination of methodologies in the study of the same
phenomenon” (1970, p. 291). Denzin distinguished between data triangulation (different
sampling strategies), investigator triangulation (different researchers to gather and
interpret data), theoretical triangulation (more than one theoretical position in interpreting
data) and methodological triangulation (using more than one method for gathering data).
The fourth type is the most common and is often used as a rationale for multi-method
research – however, the use of the term triangulation is often misunderstood (Bryman,
1984, p. 86). Greene et al. (1989) find in an empirical review on evaluation of policy
programmes that triangulation is often stated, but rarely employed in appropriate designs.
Bryman is critical towards the broad meaning of triangulation and argues that “there are
good reasons for reserving the term for those specific occasions in which researchers seek
to check the validity of their findings by cross-checking them with another method”
(Bryman, 2003, pp. 1142-1143). A broader view on triangulation comes from Rossman
and Wilson (1985), who identifies three reasons for combining quantitative and
qualitative research; 1) they can confirm or validate each other, 2) combinations can
develop analysis in order to provide richer data and 3) combinations can initiate new
modes of thinking.
This PhD project is placed within social science drawing on fields such as
consumer behaviour, social psychology and sociology exploring adolescents’
56
socialisation, social influences and possible behaviour change in relation to healthy
eating. It takes a pragmatic approach, which is quantitative dominant, according to the
continuum shown in figure 3.
Figure 3
A qualitative-quantitative continuum (Johnson et al., 2007, p. 124)
The view that triangulation is not aimed merely at validation, but at
deepening and widening one’s understanding (Greene et al., 1989; Johnson et al., 2007;
Olsen, 2004; Rossman & Wilson, 1994) is also adopted in this thesis. A pragmatic
approach in mixed methods is not easy in practical terms, though. It demands resources
from the researcher – both financially and time-wise. It takes time for the researcher to
read, familiarize with and test different research methods and especially link them in an
overall research design , but also to learn how the data gathered can be analysed using
mixed methods (Greene et al., 1989; Rossman & Wilson, 1985, 1994). The next section
will demonstrate in detail how the pragmatic approach to mixed method research was
carried out.
57
3.4. Design of empirical studies
This thesis is based on two separate, however related, empirical studies: 1) a
qualitative study consisting of (a) interviews and (b) a practical exercise with adolescents
and their families after participating in a small-scale intervention on fruit, vegetable and
sweet drinks intake, and 2) a quantitative study consisting of (a) a larger-scale
intervention study on fruit and vegetable intake and b) pre- and post-surveys. Both were
carried out as a part of the research project “Step by step changes of children’s
preferences towards healthier food” which was supported by the Danish Ministry of
Science, Technology and Innovation, grant no. 09/061357 and running from 2007 to
2013. The following gives an account for the overall research design and methods used.
3.4.1. Study 1: Qualitative study
The first study is a qualitative study that helps answer research question 1
reported in research paper 1. After having conducted a small-scale intervention study
among adolescents and their parents targeting their intake of fruit, vegetables and sweet
drinks (for example soft drinks, ice tea and drinking yoghurts), a qualitative study was
conducted in order to obtain deeper insights into adolescents’ and their families’ eating
behaviours, the adolescent-parent interaction and the different roles in the family with
regards to healthy eating. In order to understand the purpose and design of the qualitative
study, a short description of the first intervention study is sufficient. However, since the
second intervention study is a major part of this thesis’ empirical foundation – and is
based on the experiences from the first intervention study – the considerations,
reflections, and details regarding the second intervention will be thoroughly examined in
section 3.4.2.
58
In the first intervention study, 242 adolescents (age 10-12, 5th grade) were
recruited class-wise from 12 schools in the Central Denmark Region. In a nine-week
intervention period, they were randomly assigned to three experimental groups which all
received education by a dietician and filled in pre- and post-intervention surveys. Two
groups (adolescents in group 2 and adolescents and one parent in group 3) took part in a
feedback intervention using mobile phones’ Short Message Service to report on their
daily intake of fruit, vegetables and sweet drinks and receive feedback comparing the
reports to self-set weekly goals. Group 1 was a minimum intervention group. The
intervention was evaluated in terms of attrition and goal accommodation behaviours. The
sample was split in two; low vs. high intake of fruits and vegetables based on the preintervention survey. The results showed that adolescents with low pre-intervention intake
had a small, but significant increase in consumption frequency after the intervention,
whereas the partial effect on the text routines was only marginally positive (see more
details in Bech-Larsen & Grønhøj, 2013).
The qualitative study was conducted approximately three months after the
intervention and before a long-term follow-up survey, so that the families had had a
chance to reflect (and possibly forget) the intervention. The study consisted of semistructured interviews with families and a practical exercise at the families’ home. These
methods were chosen in order to gain more insight into the adolescent-parent interaction
and the socialisation regarding healthy eating, but also to explore the participants’
evaluation of the intervention including their view on the feedback and the technical side
of the intervention. According to Miles and Huberman’s four types of integrative designs
(1994) the fourth one was used in this case: Surveys were followed by a qualitative
exploration of the findings in more detail. The results of both the intervention and the
interviews were used to inform and design study 2.
59
As the purpose of the qualitative study was to obtain insights into the
evaluation of the intervention and the healthy eating socialisation in families, the goal was
to get both maximal variation in the sample – namely participants from all three
experimental groups, but also participants who were typical – meaning that they had
experienced both success and failure in the intervention. The sampling can also be viewed
as a convenience sampling with no goal of a fixed number of interviews (Flick, 2009, p.
122). If none, or only few volunteered, the strategy was to contact the families by phone
instead of a letter hoping for better recruitment. The participants in the qualitative study
were 38 families which had all taken part in the intervention and representing all 12
schools. There were ten families from the minimum intervention group, nine families
from the group, where the adolescent had reported and received feedback, and 19
families, where both the adolescent and one parent had reported and received feedback.
Since it was a qualitative study with the aim of providing understanding and insights, the
unequal distribution of families from the three groups was not seen as a problem. In all
interviews, the whole family were invited to take part which meant that the number of
interview persons ranged from two to five persons. In some cases, even divorced parents
participated together and reflected on their different experiences with the young
adolescent and healthy eating.
It was decided that the 38 interviews should be conducted by the same two
professional interviewers from a market research agency who had extensive experience
with this sort of family interviews. They were assisted by either the author of this thesis
or student assistants. Different views on the researcher’s approach to children/adolescents
are reflected within the field of child research depending on the research methods and
overall research aims (see for instance Banister & Booth, 2005). The researcher should
attempt friendship with the children/adolescents they are studying according to James et
60
al. (1998), but it is also important not to get emotionally carried away (Nørgaard, 2009).
It was therefore decided that the interviewer should be a person who had experience with
interviewing children, but did not know them. That excluded the author of this thesis,
since the author had been visiting the adolescents in school four times and instructed them
in reporting, filling in surveys and assisted in the dietician sessions.
The assistants filled out a report for each interview with background
information (see Appendix A). The general atmosphere and impression of the family and
their home was also noted paying special attention to the family members’ interaction
such as “It is a laid back family who seems to get on together well. Washed clothes yet to
be folded are lying in the sofa. Candles are lit and we are served coffee and juice. The
children are polite and sit in their seats during the whole interview – even if [the sibling]
does not say much”. Non-verbal communication such as smiles, yarning and irony was
also noted down: “The father sits most of the time with his arms crossed. [Sibling] seems
bored and yarns during the last 10 minutes of the interview”. The assistants’ observations
were cross-checked with the interviewer’s observations and hereafter reported digitally
together with the transcription of the interview.
In this sort of field work it is important that the researcher is aware of the
fact that she depends on others’ goodwill. The researcher needs to be seen as trustworthy
and able to create confidence (Douglas, 1985; Gulløv & Højlund, 2003; Nørgaard, 2009).
The interviewers therefore started with thanking the family and told them the purpose and
the duration of the interview (approx. 1.5 hour including a practical exercise midway). It
was underlined that the interviewer did not know what the family had answered in the
surveys or how they performed in the intervention. The families were granted anonymity
and could stop the interview at any time. It was underlined that there are no “right” or
“wrong” answers, and that the adolescent’s and his/her siblings’ view were as important
61
as the parents’. Consent to record the interview via a dictaphone and take a few pictures
(with the option to delete those they did not like) during the practical exercise was also
obtained before starting each interview. “[C]onfidence will usually first be earned after
more than one visit to the same family” (Nørgaard, 2009, p. 84). However, with the above
considerations and information to the family, a sense of confidence was created (Tufte,
2000) and thereby a better frame for fruitful interviews (Mayes, 2000; O'Kane, 2008).
Three interview guides reflecting the three experimental groups’
experiences in the intervention were made. They were semi-structured in order to invite
interviewees to answer using their own words and narrative structures (Flick, 2009),
thereby also taking the adolescents’ cognitive abilities into account. A common part for
all three focused on the family’s habits concerning meals, leisure time, physical activity,
grocery-shopping, creating lunch-boxes for school and cooking with emphasis on
responsibility, motivation and initiative. More indirectly, the family discourse on health
was investigated as well as its knowledge of nutrition and health recommendations. After
the practical exercise, families from the two experimental groups taking part in the report
and feedback intervention were asked to evaluate it, and especially the participating
adolescent (and possibly parent) was asked to evaluate the technical sides of the
intervention. Other family members were free to leave the interview at that point. An
example of the interview guide from the “adolescent + parent” group can be seen in
Appendix A.
When interviewing early adolescents, the interviewer must be able to
establish a sense of security, to listen, pay attention, explain and avoid giving feedback
(Banister & Booth, 2005; Gulløv & Højlund, 2003; Mayes, 2000). Young adolescents’
attitudes and behaviour can be difficult to study since their willingness to share thoughts
in an interview setting vary with their maturity, roles in the family and other
62
characteristics (shyness, little involvement etc.). Therefore, the interviewer was instructed
to make sure that all family members were heard and were given the possibility to answer
questions; as a principle, the adolescent and possible siblings were asked first and then
the parents. It is recommended that the interviewer is familiar with the age group under
study and the language they use (Banister & Booth, 2005). Hence, the interviewer was
instructed to interpret words or whole answers during the interview in order to make sure
that the statements were correctly understood. Parents sometimes answered on behalf of
the adolescent and here the interviewer asked the adolescent again. Often parents would
also supplement the adolescent’s statements, creating a dialogue. It is acknowledged that
the interview can be viewed as an intervention in itself influencing the adolescent’s and
parents’ statements.
The second qualitative method applied was a practical exercise of
approximately 20 minutes duration, where the family members had to choose what they
believed to be 600 grams of fruit and vegetable chosen from approximately five kilos of
mixed fruit and vegetables brought by the interviewer. The session was similar to the one
conducted by a dietician at school which meant that it was familiar to the adolescent, but
not to the other family members. While choosing, slicing and discussing the fruit and
vegetables, the interviewer and assistant observed the family members’ interactions and
recorded their statements. After choosing the fruit and vegetables of their choice, plates
were weighed and the difference from 600 grams was calculated making family members
reflect on this sort of illustration of six daily portions of fruits and vegetables. Besides
serving as a break in the rather long interview, the session can also be considered a
triangulation of data collection techniques which allows the researcher to both achieve
and compare different outcomes. When doing research with early adolescents, a “childcentric” approach is recommended (Banister & Booth, 2005). According to James et al.
63
engaging children in ““task-centred activities” (…) might provide a better way of
allowing children to express their ideas and opinions than the use of more “talk-centred”
methods such as interviews and questionnaires” (1998, p. 190). Banister and Booth give
an overview over innovative methods for child-centric consumer research and suggest
among other things visual aids as a way to prompt children to start talking about the
themes of interest (2005). The practical exercise is an example of a visual aid. It is a fun
element compared to the interview which might involve the risk of being taken less
serious. However, since the exercise was also for the parents and was not based on a
childish, but rather a child-centred technique (O'Kane, 2008), it was in most cases
considered fun and informative by both adolescents and parents.
When each interview was conducted, the participating assistant transcribed
the interview. Since transcription in itself is an interpretation of data (Kvale, 1997), a
transcription guide was made in order to ensure a homogeneous quality of data. Based on
a project member’s profound experience with transcribing interviews (Grønhøj, 2002)
and a method used by Silverman (2001), the guide instructed the assistants to do the
transcription verbatim without correcting the wording – even if it was not wellformulated, fluent or linguistic correct. Assistants were also instructed to mark changes in
tone of voice, laughter, pauses, hesitation, if more interview persons talked at the same
time, if things were indistinct or if the assistant had other observations (see Appendix B).
The correspondence between the recorded interviews and the transcriptions were checked
for all five assistants and was satisfactory. When writing research paper 1 (chapter 5), all
38 interviews were listened to several times and approximately 1700 pages of
transcription were assessed.
When analysing the data, the first step was to do an open coding according
to the principles of grounded theory (Strauss & Corbin, 1998). Open coding is
64
“uncovering, naming and developing concepts and categories by breaking down the data
into discrete parts, examine them closely and compare them for similarities and
differences” (Eriksson & Kovalainen, 2008, p. 164) – and is often the first step in
qualitative research (Flick, 2009). After thorough examination of the coded concepts and
categories, qualitative content analysis was carried out, where “categories are brought to
the empirical material and not necessarily developed from it” (Flick, 2009, p. 323). With
a research question to answer regarding family members’ roles in healthy eating
socialisation, useful categories were brought in from the literature on consumer
socialisation, family interaction and healthy eating and a theoretical coding was
conducted. In order to explore similarities and differences between adolescents and
parents, thematic coding was applied. One limitation to this method is that “the
categorization of text based on theories may obscure the view of the contents rather than
facilitate analysing the text in its depth and underlying meanings” (Flick, 2009, p. 328).
However, the content analysis was done without paraphrasing the text (summarizing
content analysis) which allowed a thicker description (Eriksson & Kovalainen, 2008;
Geertz, 1973) of the data.
3.4.2. Study 2: Intervention study
The second study is a quantitative study that helps answer research question
2 in research paper 2 and research question 3 in research paper 3. It is an intervention
study applying two methods: 1) An 11-week feedback intervention on fruit and vegetable
intake and 2) a pre- and post-intervention survey for adolescents and one parent.
65
3.4.2.1. Feedback intervention
Recruitment and participants
In spring 2010, 28 schools geographically distributed in the Region of
Central Denmark were approached by phone. Based on five criteria, 17 schools with a
total of 70 classes were accepted for participation. Schools were excluded if 1) there was
only one class on each level or no 9th grade, 2) if the socio-demographic background of
pupils did not more or less represent the general Danish population, 3) if there was a
lunch scheme at school (in Denmark most pupils bring their own lunch box), 4) if pupils
were allowed to leave school area and buy foods in breaks, and 5) if the school had
participated in health/food related projects within the last three years. When meeting the
criteria (six did not), school principals were informed about the research project by phone
and in more detail in a follow-up letter (survey, text message intervention, costs, 2-3
visits of one hour each). They were contacted for their consent after one week. Five
declined to participate due to time/planning. The 17 participating school principals
provided contact information for the four to six teachers at each school who were also
informed about the purpose and elements in the intervention in a letter. They all accepted
that their classes participated. Obtaining parental consent was offered, but the school
principals and teachers felt the project activities (two-three visits in school during 11
weeks and text messaging at home) could be incorporated into the “normal” school
activities, thus giving no concern for parents. Even though school principals and teachers
just wanted to be helpful, ideally consent should probably have been obtained for ethical
reasons. Before the project started, parents were informed by a similar letter as the one
sent to the teacher. Contact information for the involved researchers was provided and it
was underlined that participation in all parts of the project was voluntary and that no
compensation was offered.
66
Participants
The purpose of the intervention was to test the effect of feedback via text
messaging on adolescents’ fruit and vegetable intake (behaviour) and on important
antecedents of behaviour (self-efficacy and outcome expectations). Second, it tested
whether the combination of text messaging and a 45-minute education session on fruit
and vegetable intake by a dietician could prove more effective. The recruited schools
were therefore randomly allocated into three experimental groups. A school-wise
allocation was chosen in order not to confuse participants about the different elements;
they got the impression that the intervention was the one taking place at their school
(despite experimental groups). However, it could not be controlled whether they heard
about other experimental groups from sources outside the research project. Table 1 in
research paper 3 (Chapter 7) gives an overview over the distribution of participants in the
three groups according to class level, gender, age, parents’ education and income.
Intervention plan and elements
The intervention lasted for 11 weeks including five weeks of text messaging
for groups 2 and 3, and group 3 also received a 45-minute education from a dietician in
week 5. Group 1 was a minimum intervention group (control group). Hence, the
intervention was relatively brief which is recommended for healthy eating interventions
targeting children in order to reduce drop-out rates (Stice, Shaw, & Marti, 2006). Table 1
shows the intervention plan.
67
Table 1
Intervention plan
Week
Pre-survey
Group 3
Text messaging + nutritional
education
Pre-survey
Instruction in text messaging
Instruction in text messaging
procedure
procedure
Registration and goal-setting
Registration and goal-setting
2
Week 1: Reporting and feedback
Week 1: Reporting and feedback
3
Possibility of goal-adjustment
Possibility of goal-adjustment
4
Week 2: Reporting and feedback
Week 2: Reporting and feedback
5
Possibility of goal-adjustment
Possibility of goal-adjustment
1
Group 1
Control
group
Pre-survey
Group 2
Text messaging
45-minute class education from a
dietician
6
Week 3: Reporting and feedback
Week 3: Reporting and feedback
7
Possibility of goal-adjustment
Possibility of goal-adjustment
8
Week 4: Reporting and feedback
Week 4: Reporting and feedback
9
Possibility of goal-adjustment
Possibility of goal-adjustment
10
Week 5: Reporting and feedback
Week 5: Reporting and feedback
Post-survey
Post-survey
11
Post-survey
The participants in groups 2 and 3 were asked to report and got feedback via
text messages on their fruit and vegetable intake. In order not to exclude any participants
from the intervention, teachers were asked before the first visit to identify adolescents
who wanted to participate in the study, but did not have a mobile phone. Six participants
borrowed – by signing a contract – a mobile phone and were provided with a prepaid
mobile phone card worth 50 DKK – enough for participation during the 11 weeks.
Research assistants were instructed thoroughly before visiting the classes in
order to assure similar information was given to the participants, and each visit was
evaluated in terms of observation notes and digital sound recordings. During the first
visit, groups 2 and 3 were instructed in reporting and receiving feedback on their daily
fruit and vegetable intake via text messaging. A leaflet with six important points was
68
handed out to the participants explaining in detail about the intervention elements (see
Appendix C):
1) Participants got a unique ID-number for registration which was also
printed on both the adolescents’ and their parent’s pre- and post-surveys in order to keep
track of data.
2) Participants needed to learn how to count fruit and vegetables in “units”
(100 grams) in order to be able to set goals, report and receive feedback. Fruit and
vegetable intake was – after a pre-test with five adolescents aged 11 to 15 – considered
self-measureable for adolescents. Research assistants instructed them via a laminated A4
page with photos and verbal depictions of units equivalents of different types and
servings of fruit and vegetables, e.g. one unit could be one banana or one handful of peas
(see Appendix D). Each participant got such a page in order to support their counting
during the intervention.
3) Participants should register in the intervention by sending a text message
in class to the report-feedback-system. It was again underlined that participation was
voluntary, and the costs for text messages (approximately 10 DKK) would not be
refunded. Participants immediately got a text message back from the system asking for a
weekly goal for fruit and vegetable intake in units. Following the recommendations of
Kluger and Denisi (1996) and Locke et al. (1981), the participants were asked to set
ambitious, yet realistic goals for their weekly fruit and vegetable intake. In order to
maximize adherence and performance, these goals could be adjusted every other week
(cf. section 2.4).
69
4) An example of the request-report-feedback loop was illustrated, and
participants were instructed how to avoid error messages by sending reports and goals in
the correct form.
5) A Q&A section for help in case of problems (no/wrong/forgot text, lost
phone, change of goals etc.) and information about how to reach a hotline via text
messages every day until 11 pm and a project e-mail address for questions.
6) A prize (a trip to a local fun park worth approx. 525 Euro) was promised
to the class sending most of the possible text messages. According to Raju et al. (2010)
competitions in interventions can increase socially desirable responding, but this type of
effect was expected to be lower when making the prize dependent on collective rather
than individual responses, and when promoting participation rather than increased intake
of fruit and vegetables.
7) A time plan (the same given to parents) was presented in order to keep
track of the intervention.
By providing the adolescents with detailed information about the research
project, they were deemed capable of giving “informed consent” (France, 2004) –
especially when non-participation (no registration or no reports) was an option.
Setup of feedback system
A two-way text messaging approach consisting of reporting and feedback
was automated by a web-based software program specifically developed for the
intervention. Participants were made aware that the registration of consumption and goal
data was fully automatic and anonymous (only for use in this research project). Every
70
night at 8 pm in the weeks of reporting, participants were requested to send a report
covering that day’s units (for instance 2 f, 3 v). Participants immediately received
feedback from the system comparing their daily report to the self-set weekly goal. This
should provide the adolescents with a sense of reality; if they were short of eating for
instance 14 fruits Saturday (with the registration ending Sunday), they should either come
to the conclusion that the weekly goal was too ambitious or that their efforts were not
spread out evenly during the week.
Previous studies using text messaging have shown that framing of messages
(Woolford, Clark, Strecher, & Resnicow, 2010), message frequency and timing (Haug,
Meyer, Schorr, Bauer, & John, 2009; Head, Noar, Iannarino, & Grant Harrington, 2013)
influence participation as well as outcomes. Schultz and colleagues (2007) did a feedback
study on household energy conservation using normative messages. A descriptive
normative message compared a household’s energy usage to the neighbourhood average
produced either desirable energy savings or a boomerang effect (depending on low or
high baseline rate). When adding an injunctive message in the form of an emoticon
conveying social approval or disapproval, the boomerang effect was eliminated. Including
normative messages in feedback studies is interesting, but when the aim is to improve
adolescents’ fruit and vegetable intake, self-efficacy and outcome expectations, normative
messages could blur the picture, as adolescents can be especially sensitive to peerpressure. Hence, the feedback was factual and did just calculate differences between goal
and intake. It was framed in concrete terms in order to secure higher adherence and
performance levels (Bech-Larsen & Grønhøj, 2013; Kluger & DeNisi, 1996). Figure 4
displays participants’ registration, goal-setting and the daily request-report-feedback loop.
71
Figure 4
Registration and goal-setting and daily request-report-feedback loop (participants in green boxes, system in
white)
Registration text to system
Confirmation and request
for first week’s goals
Reporting first week’s
goals
Daily request
Daily report
Daily feedback
“Please reply with the number of
units of fruit and vegetables
consumed today. Your text
should look like this: 1f, 2v”
“2f, 3v”
“You have now eaten 7f and
6v. In order to reach your
weekly goal you still need 14f
and 15v before Sunday
evening”
A software programmer was hired to create a system which could keep track
of all participants, their goals, goal-adjustments, reports and feedback. The setup for
registration and first goal is illustrated in figure 5, and the setup for the daily requestreport-feedback loop is illustrated in figure 6.
72
Figure 5
Registration and goal-setting (system side)
Start
Wait for
incoming text
”1f, 2v”
Interpret the text as
a new goal
Yes
Registered
phone
number?
known?
No
Interpret the
text as a
registration
Registered
ID number?
New goal
OK?
Yes
Yes
No
Connect sender’s
phone number with ID
number.
Reply with confirmation
of new goal
Send error message
Wait for incoming text
73
Reply ”registration
done” and ask for first
week’s goal
Figure 6
Change of goal, registration and daily request-report-feedback loop (system side)
Send text: ”A new
week begins….change
of goal?”
Start
6 pm
Send text: ”Your goal
is now…” or “error”
Wait for incoming
report: “1f, 2v”
6 pm
Save the new goal
Go to next
reporting
no
Is the week
over?
yes
Stop
Send error message
to participant
Send daily request
Wait for incoming
report: “1f, 2v”
Interpret
report
succes
spm
Send reminder to all,
who have not sent
daily report
Send feedback (goalreport comparison)
File report under
actual registration
day (overwrite if
already registered
data)
Response rates and quality of data
Both setups were tested first with all researchers in the project including
research assistants for a week and after making some adjustments, two pre-tests were run
with seven adolescents (11-15 years) for a week. In the intervention, the total drop-out
from text message-reporting ranged from 16.6% in week 1 to 49.2% in week 5 (see figure
1 in research paper 3, chapter 7). Very few community-based health intervention studies
targeting adolescents report drop-out rates. However, since this feedback intervention
aimed at developing a realistic tool for health promotors (in terms of costs and time), a
setup as close as possible to real life was chosen – with a high drop-out rate as a possible
consequence.
74
In the analysis phase it was clear that the text message data was not
completely valid. Even though the feedback system did not allow reports stating more
than 10 units of fruit and vegetables, respectively, per day, some participants had reported
amounts exceeding 20 units per day – often in several consecutive days indicating errors
in the reports. Weekly averages were calculated since there was a pattern indicating that
participants seemed to forget to report one day and then added two reports together on the
next. Here, it could have been useful to have further control measures to ensure
reliability.
Nutritional education for group 3
In order to test whether the text messaging effects would be stronger when
combined with nutrition education by a dietician, classes in group 3 were visited by a
dietician from Centre for Public Health, Central Denmark Region. Following the official
health recommendations (6 portions of fruit and vegetables per day), the dietician
educated (in a 45-minute session) the adolescents about the importance of eating fruit and
vegetables. A practical exercise (the one also conducted in the interviews described in
section 3.4.1.) was conducted where participants in groups were asked to choose what
they believed to be 600 grams from five kilos of mixed fruits and vegetables. The
amounts were later weighted and the difference to 600 grams was calculated, thereby
visualising both the estimation and the recommended 600 grams of fruit and vegetables
for the participants. Dieticians also handed out a flyer to the participants to bring home
(see Appendix E). The content of the education sessions was coordinated between two
dieticians and five research assistants in order to ensure that all classes received identical
information. This was confirmed by audio recordings and observations of the sessions.
75
3.4.2.2. Pre-and post-intervention surveys
The purpose of the surveys was to measure intervention effects. Identical
surveys (despite socio-demographic measures in the baseline survey) were handed out to
participants in all three groups and to one of their parents. According to the literature on
child research, surveys for parents can (with care) also be used for children aged 11+
(Christensen & James, 2008). However, it is a challenge to develop surveys targeting both
adolescents and parents and therefore, the survey pre-tested among adolescents was used.
Where appropriate a 5-point Likert scale ranging from 1 = totally disagree to 5 = totally
agree was used, since children (6-18) reportedly prefer the Likert scale over other scales
(van Laerhoven, van der Zaag-Loonen, & Derkx, 2004), and five points because children
are often unable to grasp seven or more response options (Chambers & Johnston, 2002;
Streiner & Norman, 2008).
Adolescents answered the surveys in class supervised by research assistants,
and at home the parents did the same instructed by a letter and provided with a return
envelope. The advantages of answering the surveys in class were the high response rates
(88.5 % and 88.2 %, for pre- and post-surveys, respectively) and that research assistants
could supply further information and answer questions. The latter could also prove as a
disadvantage, since participants’ answers could be influenced. Also, a possible
competition in class to finish first could influence the answers. As for parents, response
rates were rather low (52.8 % and 25.1 %, respectively) which might reflect the more
uncontrolled setting at home. The ID-number also used in the intervention linked
adolescents’ and parents’ answers in both the pre- and post-survey.
For pre-testing the survey, a school gave access to four adolescents (age 11).
They were labelled slow, medium and fast readers by their teacher. They answered the
survey and discussed afterwards wording and scales. Following adjustments, four new
76
adolescents from the same school went through the same pre-test. A third pre-test was
conducted with 30 adolescents (age 10-16) who answered the survey and subsequently
frequencies and scale reliabilities were checked. The pre-tests indicated that 30 minutes
was needed to answer the survey, but during the visits in class, some participants needed
more time and they were offered the possibility of finishing it at another time in school
under the teacher’s supervision. The survey was also pre-tested by 10 adults.
When doing intervention studies, it is recommended to do a long-term
follow up in order to check the strength of possible behaviour changes (Stice et al., 2006)
(as done in the pilot study described in section 3.4.1.). However, due to funding issues,
this was not possible in the present case.
Measures
The measures used in the main analysis were behaviour (frequency of intake
of fruit and vegetables for breakfast, lunch, dinner and between meals during a week,
eight items), self-efficacy, outcome expectations and perception of others’ behaviour
(Bandura, 1977). The scales and items are explained in detail in research paper 2 (Chapter
6) and 3 (Chapter 7). Other measures in the survey, but not included in the main analyses,
were: mastery experience (Bandura, 2004), social-structural factors (Bandura, 2005),
motivation for health behaviours (autonomous motivation, controlled motivation and
amotivation (Deci & Ryan, 1985)), general self-efficacy (Luszczynska, Scholz, &
Schwarzer, 2005; Schwarzer & Jerusalem, 1995) and intentions to eat healthier (Bandura,
1997).
77
Figure 7 gives an overview over the research design applied in this thesis. It
summarizes the empirical studies described in section 3.4. and shows how they relate to
the three research papers.
Figure 7
Research design
Research paper 1
Research paper 2
Research paper 3
Qualitative methods
Quantitative method
cross-sectional
Quantitative method
longitudinal
In-depth
interviews with
families
n = 38
Practical
exercise with
families
n = 38
Preintervention
survey
adolescentparent dyads
n = 757
78
Pre- and postintervention
surveys,
adolescents
n = 1488
Feedback
intervention
with
adolescents for
11 weeks
n = 986
References
Andersen, D. (2000). Spørgeskemainterview med børn - skulle det være noget særligt?
[Survey interviews with children - is that something special?]. In P. S. K.
Jørgensen, J. (Ed.), Børn som informationer - antologi (pp. 71-84). Copenhagen:
Børnerådet [The National Council for Children].
Andersen, D., & Kjærulff, A. (2003). Hvad kan børn svare på? - Om børn som
respondenter i kvantitative spørgeskemaundersøgelser [What can children
answer? - Children as respondents in quantitative surveys]. Copenhagen: SFI The Danish National Centre for Social Research.
Andersen, D., & Ottosen, M. H. (2002). Børn som respondenter. Om børns medvirken i
survey. [Children as respondents. On children's participation in surveys].
Copenhagen: The Danish National Centre for Social Research (SFI).
Bandura, Albert. (1977). Social learning theory (1st ed.). Englewood Cliffs, NJ: PrenticeHall.
Bandura, Albert. (1997). Self-efficacy: The exercise of control. New York: W. H.
Freeman.
Bandura, Albert. (2004). Health promotion by social cognitive means. Health Education
& Behavior 31(2), 143-164.
Bandura, Albert. (2005). The evolution of social cognitive theory. In K. G. H. Smith, M.
A. (Ed.), Great minds in management (pp. 9-35). Oxford: Oxford University
Press.
Banister, Emma N., & Booth, Gayle J. (2005). Exploring innovative methodologies for
child‐centric consumer research. Qualitative Market Research: An International
Journal, 8(2), 157-175. doi: doi:10.1108/13522750510592436
Bech-Larsen, Tino, & Grønhøj, Alice. (2013). Promoting healthy eating to children: A
text message (SMS) feedback approach. International Journal of Consumer
Studies, 37(3), 250-256. doi: 10.1111/j.1470-6431.2012.01133.x
Bryman, Alan. (1984). The debate about quantitative and qualitative research: A question
of method or epistemology? The British Journal of Sociology, 35(1), 75-92. doi:
10.2307/590553
Bryman, Alan. (2003). Triangulation. In M. Lewis-Beck, Bryman, Alan E, Liao, Tim
Futing (Ed.), Encyclopedia of Social Science Research Methods (pp. 1142-1143).
London: SAGE Publications.
79
Bryman, Alan. (2006). Paradigm peace and the implications for quality. International
Journal of Social Research Methodology, 9(2), 111-126. doi:
10.1080/13645570600595280
Christensen, P., & James, A. (2008). Research with children: Perspectives and practices.
London: Falmer Press.
Creswell, John W. (2011). Controversies in mixed methods research. The Sage handbook
of qualitative research, 4, 269-284.
Creswell, John W., & Plano Clark, Vicki L. (2011). Designing and conducting mixed
methods research (2nd ed.). Los Angeles: SAGE Publications.
Croll, J. K. , Neumark-Sztainer, D., & Story, M. (2001). Healthy eating. What does it
mean to adolescents? Journal of Nutrition Education (33), 193-198.
Deci, E, & Ryan, R. (1985). Intrinsic motivation and self-determination in human
behavior. New York: Plenum Press.
Denzin, N. K. (1970). The research act: A theoretical introduction to sociological
methods. New York: Aldine.
Denzin, N. K. (2009). The elephant in the living room: Or extending the conversation
about the politics of evidence. Qualitative Research, 9(2), 139-160. doi:
10.1177/1468794108098034
Douglas, Jack D. (1985). Creative interviewing. Beverly Hills: Sage.
Eriksson, P., & Kovalainen, A. (2008). Qualitative methods in business research.
London: Sage Publications.
Falk, Laura Winter, Sobal, Jeffery, Bisogni, Carole A., Connors, Margaret, & Devine,
Carol M. (2001). Managing healthy eating: Definitions, classifications and
strategies. Health Education & Behavior, 28(4), 425-439. doi:
10.1177/109019810102800405
Flick, U. (2009). An introduction to qualitative research (4th ed.). London: Sage
Publications.
France, Alan. (2004). Young people. In S. Fraser, V. Lewis, S. Ding, M. Kellett & C.
Robinson (Eds.), Doing research with children and young people (pp. 175-190).
London: SAGE Publications.
Geertz, Clifford. (1973). Thick description: Toward an interpretive theory of culture. In
C. Geertz (Ed.), The interpretation of cultures: Selected essays. New York: Basic
Books.
80
Greene, Jennifer C., Caracelli, Valerie J., & Graham, Wendy F. (1989). Toward a
conceptual framework for mixed-method evaluation designs. Educational
Evaluation and Policy Analysis, 11(3), 255-274. doi:
10.3102/01623737011003255
Grønhøj, Alice. (2002). Miljøvenlig adfærd i familien: Et studie af familiemedlemmernes
involvering og sociale interaktion [Pro-environmental behaviour in the family: A
study of family members' involvement and social interaction]. Århus: Institut for
Markedsøkonomi.
Gulløv, Eva, & Højlund, Susanne (2003). Feltarbejde blandt børn: Metodologi og etik i
etnografisk børneforskning [Field research among children: Methodology and
ethics in ethnographic research with children] (1st ed.). Copenhagen: Gyldendal.
Hall, Ralph (2012). Mixed methods: In search of a paradigm. Retrieved January 7,
2015, from www.auamii.com/proceedings_phuket_2012/hall.pdf
Haug, Severin, Meyer, Christian, Schorr, Gudrun, Bauer, S, & John, Ulrich. (2009).
Continuous individual support of smoking cessation using text messaging: A pilot
experimental study. Nicotine & tobacco research: Official Journal of the Society
for Research on Nicotine and Tobacco, 11(8), 915-923.
Head, Katharine J, Noar, Seth M, Iannarino, Nicholas T, & Grant Harrington, Nancy.
(2013). Efficacy of text messaging-based interventions for health promotion: A
meta-analysis. Social Science & Medicine (1982), 97, 41-48.
James, Allison, Jenks, Chris, & Prout, Alan. (1998). Theorizing childhood. London:
Polity Press.
John, Deborah Roedder. (1999). Consumer socialization of children: A retrospective look
at twenty-five years of research. Journal of Consumer Research, 26(3), 183-213.
Johnson, R. Burke, & Onwuegbuzie, Anthony J. (2004). Mixed methods research: A
research paradigm whose time has come. Educational Researcher, 33(7), 14-26.
doi: 10.3102/0013189x033007014
Johnson, R. Burke, Onwuegbuzie, Anthony J., & Turner, Lisa A. (2007). Toward a
definition of mixed methods research. Journal of Mixed Methods Research, 1(2),
112-133. doi: 10.1177/1558689806298224
Kampmann, J. (1998). Børneperspektiv og børn som informanter [Childrens' perspective
and children as informants]. Copenhagen: The National Council for Children.
Kazbare, L. (2010). Healthy eating behaviour - a social marketing perspective. (PhD
Thesis), MAPP Centre, Aarhus School of Business, Aarhus University. Retrieved
from https://pure.au.dk/ws/files/14287/Kazbare_2010
81
Klesges, Lisa M., Baranowski, Tom, Beech, Bettina, Cullen, Karen, Murray, David M.,
Rochon, Jim, & Pratt, Charlotte. (2004). Social desirability bias in self-reported
dietary, physical activity and weight concerns measures in 8- to 10-year-old
African-American girls: Results from the Girls health Enrichment Multisite
Studies (GEMS). Preventive medicine, 38, Supplement(0), 78-87. doi:
10.1016/j.ypmed.2003.07.003
Kluger, A.N., & DeNisi, A. (1996). The effects of feedback interventions on
performance: A historical review, a meta-analysis and a preliminary feedback
intervention theory. Psychological Bulletin, 119(2), 254-284.
Kvale, Steinar. (1997). Interview: En introduktion til det kvalitative forskningsinterview.
København: Hans Reitzels Forlag.
Lake, A. A., Hyland, Robert M., Rugg-Gunn, Andrew J., Wood, Charlotte E., Mathers,
John C., & Adamson, Ashley J. (2007). Healthy eating: Perceptions and practice
(the ASH30 study). Appetite, 48(2), 176-182.
Locke, Edwin A, Shaw, Karyll N, Saari, Lise M, & Latham, Gary P. (1981). Goal setting
and task performance: 1969-1980. Psychological bulletin, 90(1), 125-152. doi:
10.1037/0033-2909.90.1.125
Luszczynska, A., Scholz, U., & Schwarzer, R. (2005). The general self-efficacy scale:
Multicultural validation studies. The Journal of Psychology, 139(5), 439-457.
Margetts, B. M., Martinez, J. A., Saba, A., Holm, L., & Kearney, M. (1997). Definitions
of 'healthy' eating: A pan-EU survey of consumer attitudes to food, nutrition and
health. European Journal of Clinical Nutrition, 51, 23-29.
Mayes, P. A. (2000). Designing questions for children. In J. Y. Le Bigot (Ed.),
Researching youth (pp. 25-39). Amsterdam: ESOMAR.
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded
sourcebook (2nd ed.). Newbury Park, CA: Sage.
Morgan, David L. (2007). Paradigms lost and pragmatism regained: Methodological
implications of combining qualitative and quantitative methods. Journal of Mixed
Methods Research, 1(1), 48-76. doi: 10.1177/2345678906292462
Nørgaard, Maria Kümpel. (2009). Family decision-making during food buying. (PhD
thesis), MAPP Centre, Aarhus School of Business, Aarhus University.
O'Kane, C. (2008). The development of participatory techniques - Facilitating children's
views about decisions which affect them. In P. Christensen & A. James (Eds.),
Research with children - Perspectives and practices (pp. 136-159). London:
Falmer Press.
82
Olsen, W. (2004). Triangulation in social research: Qualitative and quantitative methods
can really be mixed In M. Holborn (Ed.), Developments in Sociology. Ormskirk:
Causeway Press.
Povey, R., Conner, M., Sparks, Paul, James, R., & Shepherd, Richard. (1998).
Interpretations of healthy and unhealthy eating and implications for dietary
change. Health Education Research, 13(2), 171-183.
Raju, S., Rajagopal, P., & Gilbride, T. (2010). Marketing healthful eating to children: The
effectiveness of incentives, pledges and competitions. Journal of Marketing, 74,
93-106.
Rossman, Gretchen B., & Wilson, Bruce L. (1985). Numbers and words: Combining
quantitative and qualitative methods in a single large-scale evaluation study.
Evaluation Review, 9(5), 627-643. doi: 10.1177/0193841x8500900505
Rossman, Gretchen B., & Wilson, Bruce L. (1994). Numbers and words revisited: Being
“shamelessly eclectic”. Quality and Quantity, 28(3), 315-327. doi:
10.1007/bf01098947
Schultz, P. Wesley, Nolan, J. M., Cialdini, R. B. , Goldstein, N. J. , & Griskevicius, V.
(2007). The constructive, destructive and reconstructive power of social norms.
Psychological Science, 18(429). doi: 10.1111/j.1467-9280.2007.01917.x
Schwarzer, R., & Jerusalem, M. (1995). Generalized self-efficacy scale. In J. Weinman,
S. Wright & M. Johnston (Eds.), Measures in health psychology: A user’s
portfolio. Causal and control beliefs (pp. 35-37). Windsor, England: NFERNELSON.
Scott, J. (2008). Children as respondents - The challenge for quantitative methods. In P. J.
Christensen, A. (Ed.), Research with children - Perspectives and practices (pp.
87-108). London: Falmer Press.
Silverman, David. (2001). Interpreting qualitative data: Methods for analysing talk, text
and interaction (2. ed. ed.). London: SAGE.
Stice, Eric, Shaw, Heather, & Marti, C. Nathan. (2006). A meta-analytic review of
obesity prevention programs for children and adolescents: The skinny on
interventions that work. Psychological Bulletin, 132(5), 667-692.
Strauss, Anselm;, & Corbin, Juliet. (1998). Basics of qualitative research: Techniques
and procedures for developing grounded theory (Second edition ed.). Thousand
Oaks, CA: Sage.
Tashakkori, Abbas, & Teddlie, Charles. (2010). SAGE handbook of mixed methods in
social & behavioral research (2nd ed.). Thousand Oaks, Calif.: SAGE.
83
Thomas, J., Sutcliffe, K., Harden, A., Oakley, A., Oliver, S., Rees, R., . . . Kavanagh, J.
(2003). Children and healthy eating: A systematic review of barriers and
facilitators. London: EPPI-Centre, Social Science Research Unit, Institute of
Education, University of London.
Tufte, B. (2000). At forske for og med børn - hvilke etiske udfordringer stiller intentionen
om at bruge børn som informanter? [To do research for and with children - which
ethical challenges are connected to the intention about using children as
informants?]. In P. S. Jørgensen & J. Kampmann (Eds.), Børn som informanter Antologi [Children as informants - An antology]. Copenhagen: The National
Council for Children.
Woolford, Susan J, Clark, Sarah J, Strecher, Victor J, & Resnicow, Kenneth. (2010).
Tailored mobile phone text messages as an adjunct to obesity treatment for
adolescents. Journal of Telemedicine and Telecare, 16(8), 458-461. doi:
10.1258/jtt.2010.100207
84
Appendixes A-E
Appendix A. Observation sheet and interview guide (in Danish)
Observationer af familieinterview
Dato for interview
Tidspunkt
Løbenummer (se interviewliste)
Gruppe (se interviewliste)
Interviewer
Assistant
Hvilken forælder har udfyldt
spørgeskemaet?
Hvilken forælder har rapporteret via sms?
(KUN GRP. 2)
Deltagere i interview (sæt x og udfyld)
 Deltagende elev
 Søskende 1
 Søskende 2
 Mor
 Far
 Andre (angiv)
Afvejningsseance
Antal gram
/
Start:
Slut:
 Mor
 Far
 Mor
 Far
Navn
Alder
Ca. tid for at finde
frugt og grønt
Ca. grøntsags-/frugt ratio (fx
½, ¼ g og ¾ f)
 Deltagende elev
 Søskende 1
 Søskende 2
 Mor
 Far
 Andre (angiv)
Antal billeder taget
Indtryk af familien:
Observeret non-verbal kommunikation (fx smil, ironi, kedsommelighed etc.):
Indtryk af stemningen under interviewet (fx trykket, humoristisk etc.):
85
Spørgeguide, gruppe 2
Interview guide
Blå markeringer er særligt for interviewer
1. Introduktion (5-10 min)
Først og fremmet vil vi sige tak for, at I har sagt ja til at deltage, og at vi måtte komme her og besøge jer.
Jeg vil interviewe jer, mens min assistent [navn] vil sørge for, at teknikken fungerer og notere vigtige ting
ned.
Jeg vil først fortælle lidt om, hvor vi befinder os i projektet, så vil jeg fortælle lidt om interviewet og hvad
vi bruger det til. Men allerførst vil jeg gerne vide, hvem af jer [far eller mor], der har udfyldt
spørgeskemaerne:
 Mor
 Far
Og hvem af jer har deltaget i sms-rapporteringen:
 Mor
 Far
Jeg vil også gerne vide, hvad I hedder, og hvor gamle I er:
Person
Navn
Alder
Deltagende elev
Søskende 1
Søskende 2
Mor
Far
1) Om projektet: I har jo deltaget i Step by Step projektet ved at udfylde spørgeskemaer, og du [barnet] og
du [forælder] har rapporteret vha. sms. Nu taler vi så med en række familier i projektet, for at få uddybet de
resultater, der blev indsamlet gennem de to første spørgeskemaer. Jeg vil gerne understrege, at jeg ikke ved,
hvad I har svaret i spørgeskemaerne. I slutningen af oktober kommer så det tredje og sidste spørgeskema.
2) Om interviewet: Interviewet kommer til at vare ca. 1½ time, og jeg vil først spørge lidt ind til jeres vaner
i forhold til mad. Så skal vi arbejde lidt med noget frugt og grønt, vi har medbragt, mens vi slutter af med at
spørge ind til jeres [barnet og den deltagende forælders] oplevelse af at være med i projektet. Her behøver I
andre ikke at sidde med ved bordet.
Det er vigtigt for mig at understrege, at der ikke er ”rigtige” eller ”forkerte” svar. Vi er interesseret i det, I
gør. Altså jeres egne erfaringer og oplevelser – ikke bare jeres holdninger. Det er vigtigt for os at høre om
jeres erfaringer hver især og alles erfaringer er lige vigtige, så I skal endelig ikke holde jer tilbage.
Interviewet bliver optaget, så vi kan lave en udskrift af interviewet, som vi kan bruge senere hen, når vi skal
analysere resultaterne.
Intervieweren skal sikre, at alle tidligt kommer til orde – og huske at nævne familiens navne naturligt før
eller i forlængelse af deres svar.
86
Formålet er at skabe et grundlag for at forbedre ernæringsoplysningen i folkeskolen, og at skrive nogle
videnskabelige artikler, men alt hvad I har sagt, bliver selvfølgelig behandlet anonymt.
Har I nogen spørgsmål? Hvis ikke, så lad os komme i gang.
2. Fokus på familiens vaner, roller og interaktion: Generelt og projektrelateret (20-25 min)
Nu vil jeg gerne høre lidt mere om, hvilke vaner I har i jeres familie, når det gælder mad.
Morgenmaden
o
Spiser I alle morgenmad?
 Hvis nej, hvorfor ikke?
o Hvad spiser I til morgen?
 Børnene spørges direkte først
o Hvem bestemmer, hvad der spises til morgenmad?
 Henvendt til børnene, har I selv bestemt, hvad I vil spise til morgenmad?
o Har I ændret på den morgenmad I spiser, efter at I er kommet med i dette projekt?
 Hvis ja, hvordan kan det være?
Madpakke
o
Har I madpakke med i skole/på arbejde?
 Start med børnene – spørg direkte
o Hvordan foregår det, når I laver madpakker?
 Hvem står for at lave madpakkerne?
 Hjælper I hinanden eller er der altid en, der sørger for det?
 Børnene direkte – Er I med til at bestemme hvad, der skal i madpakken?
 Hvis nej, kan I lide det?
 Har madpakkerne ændret sig de sidste fire måneder?
 Er det noget andet I får med på madpakken nu?
 Hvis ja, spørges børnene om hvor tilfredse de er med madpakken nu?
Aftensmad
o
Beskriv hvordan det typisk foregår, når I laver aftensmad.
 Hvem bestemmer, hvad I skal spise til aftensmad?
 Laver I madplaner?
 Børnene spørges direkte: Har I en ”mad-dag”/pligter i forbindelse med madlavning og indkøb?
 Har aftensmaden ændret sig de sidste fire måneder? Børnene spørges direkte først
 I hvilken grad har I mærket ændringerne?
 Hvem har sørget for det?
 Hvad synes I andre om de ændringer?
o
Har I talt om det/ændringerne? som familie?
 Hvad har I så talt om? / Hvordan taler I om det?
 Hvem af jer har taget det op/tager initiativet (far/mor, barn)?
Frugt og grønt
o
o
o
Hvad synes I om at spise frugt og grønt? (kan de godt lide de, eller er det noget, de gør af pligt)
 Hvilke slags foretrækker I? Hvad har I fx i huset nu? (børn spørges først)
Gør I noget særligt for at spise nok frugt og grønt?
 Hvad?
 Hvorfor/hvorfor ikke?
 Hvem sørger for det?
Synes I, at det med at spise frugt og grønt har ændret sig, efter I har deltaget i projektet?
 Hvis ja, på hvilken måde?
87


Hvem har sørget for det?
Hvad synes I andre om de ændringer? Børnene spørges direkte først
Søde drikke
o
o
o
o
o
Hvilke søde drikke (kakaomælk, drikkeyoghurt, iste, sodavand, saftevand etc.), drikker I? (børnene
spørges direkte først)
Hvornår drikker I søde drikke?
 Børnene spørges direkte
Har I som familie nogle regler for at drikke søde drikke (fx kun i weekenden, kun en vis mængde pr.
dag/uge etc.)?
 Hvilke?
 Hvem bestemmer?
 Bliver de overholdt af alle i familien?
Synes I, at det har ændret sig, efter I har deltaget i projektet?
 Hvis ja, på hvilken måde har det ændret sig?
 Hvem har sørget for det?
 Hvad synes I andre om de ændringer? Børnene spørges direkte først
Har I talt om det som familie?
 Hvad har I talt om?
Mellemmåltider
Nu har vi talt om de tre hovedmåltider. Hvad så med mellemmåltider? (børnene spørges direkte først)
o
o
o
Hvad spiser I til mellemmåltid?
Bestemmer I selv, hvad I vil spise til mellemmåltid?
Har I det med hjemmefra eller er det noget I køber?
Indkøb (nedprioriteres hvis presset på tid)
Vi har nu talt om de forskellige måltider i løbet af dagen. Jeg vil gerne ganske kort runde indkøbene til alle
måltiderne.
o
Hvordan foregår indkøbene?
 Hvem er det, der beslutter hvad der skal købes i løbet af ugen?
 Børnene: er I også med til at bestemme?
 Hvem køber ind?
Lidt om fritid:
Til børnene:
o
Hvad laver I, når I ikke er i skole?
 Går I til aktiviteter i fritiden? Hvilke?
 Cykler eller spiller fodbold med vennerne i fritiden? – Aktivt fritidsliv med bevægelse eller
stillesiddende foran computer?
 Hvad laver I i frikvartererne på skolen?
 Hvis leg, hvilken type leg?
Til de voksne:
o
Dyrker I motion?
 Hvad?
 Hvor tit?
88
o
Gør I noget i det daglige for at røre jer (fx gåture, cykle til arbejde, tage trappen etc.)?
o
Er det vigtigt for jer at dyrke motion?
 Hvorfor/hvorfor ikke?
Har I ændret nogen ting vedr. motion, efter I har deltaget i projektet?
 Hvis ja, hvad har I så ændret?
Har I talt om det i familien?
o
o
Øvrigt
o
Har I erfaringer med at gøre noget særligt for at spise sundere?
o
Lægger I mere mærke til information om sundhed og ernæring, fx artikler, kampagner m.v., efter I har
deltaget i projektet? Direkte til børnene først og derefter forældre
o
Kan I huske slogans og anbefalinger om sundhed og ernæring? Hvilke?
3. Afvejningsseance (15-20 min)
Der skal bruges:
o
o
o
o
o
o
Skærebrædt (medbringes)
Kniv (medbringes)
Vægt (medbringes)
Voksdug (medbringes)
En kasse frugt og grønt (medbringes)
Kamera (medbringes)– så assistenten kan tage billeder af afvejningsseancen
Nu vil jeg bede jer om at se på det frugt og grønt, jeg har medbragt. Du (barnet) kender det fra diætistens
besøg. Det, I skal gøre nu, er hver at forsøge at finde 600 gram frugt og grønt, som I kunne tænke jer at
spise i løbet af en dag. I må gerne skære frugten og grøntsagerne i stykker, og så lægger I det, I vælger, på
jeres tallerken, som vi så vejer til sidst.
Vi vil meget gerne have lov at fotografere jer, mens I arbejder med det medbragte frugt og grønt. Billederne
vil kun blive brugt, når undersøgelsens resultater skal præsenteres. Hvis I ønsker det, kan vi kigge
billederne igennem bagefter og slette dem, I ikke kan lide.
Alle kan tage frugt og grønt, men de skal gøre det uden at kigge alt for meget på hvad de andre tager.
Assistenten registrerer, hvor mange gram frugt og grønt, hvert familiemedlem har valgt:
Person
Antal gram
Ca. tid for at finde frugt og grønt
Ca. grøntsags/frugt ratio
Deltagende elev
Søskende 1
Målet er at
ramme 600
g (svarer til
”6 om
dagen”)
Søskende 2
Mor
Far
Prøv at
89
fortæl lidt om det, I hver især har valgt (start med børnene).
o
o
o
o
Svarer det til den mængde og de typer af frugt og grønt, som I spiser på en dag?
Spiser I mere/mindre?
Forsøger I at nå ”6 om dagen”?
Er I overrasket over jeres ”resultat” (afvejningen)?
4. Familiens oplevelse af projektet (10 -15 min)
Nu vil jeg gerne høre mere om, hvordan I som familie har oplevet Step by Step projektet:
Overordnet
o
o
Hvordan har det været at være med i projektet?
Har I alle sammen syntes, at det har været xx (deltagernes ord – sjovt, interessant eller hvad det nu må
være)?
 Hvad har gjort størst indtryk? Er der nogen ting, der har gjort større indtryk end andre?
 Hvad har fungeret godt/skidt? Er der nogen ting I synes har fungeret bedre end andre?
 Hvad har I talt om i familien?
Tælleskema
o
Hvad synes I om tælleskemaet?
 Har I talt om eller brugt tælleskemaet? [fremvis hvis familien ikke har det]
 Bruger I det evt. stadig?
 Hvis nej, hvorfor ikke?
Diætistbesøg
o
Hvad synes du (eleven) om at have diætisten på besøg i klassen?
 Talte I om besøgene og projektet i klassen?
 Hvad talte I om?
 Er det et emne som I har arbejdet videre med i klassen?
 Talte I om besøgene hjemme?
 Hvad talte I om?
 Hvad har det betydet for resten af familien, at X har fået den vejledning fra diætisten
og har fået materialer med hjem?
 Har I brugt det materiale, som X fik med hjem fra diætisten?
 Hvordan har I brugt det?
 Har Xs deltagelse i sms-rapporteringen fået nogle af jer andre til at ændre eller tænke
mere over jeres spisevaner?
 Hvordan?/hvorfor?
 Kunne I andre tænke jer at have været mere direkte inddraget i forløbet?
 Hvis ja, på hvilken måde? (fx også modtage sms’er og kostvejledning?)
Spørgeskema
o
Nu har det været Y (far/mor), der har udfyldt spørgeskemaet. Hvordan kom I frem til, at det var
dig/ham/hende?
 Kunne du (far/mor) ligeså godt have gjort det?
Snak med andre
90
o
I har fortalt om, hvordan I her i familien har snakket om de forskellige aktiviteter i projektet og
ændringer. Har I snakket med andre om det?
 Hvem har I talt med om det?
 Hvad har I snakket om?
 Synes I i det hele taget, at I efter at have været med i projektet er begyndt at snakke mere med
jeres venner og kollegaer om kost og motion?
Ændringer
o
Tror I, at I som familie vil ændre nogle ting omkring indkøb og spisning fremover?
 Hvorfor? / Hvad? (fx antal måltider, sundere mad, fx købe ind andre stedet etc.)?
 Hvorfor ikke? Hvad holder jer tilbage?
o
Tror I, at I som familie vil ændre nogle ting omkring motion fremover?
 Hvorfor? Hvad?
 Hvorfor ikke? Hvad holder jer tilbage? (Referer til de ændringer, de formentlig allerede har
omtalt under punkt 1)
o Hvis ja: Hvis jeg nu kommer igen om en måned, holder I så stadig fast i ændringerne?
 Kost
 Indkøb
 Motion (provoker familien, for det er nemt at sige, at man vil ændre noget)
o Hvis nej, hvad tror I der skal til for, at I vil holde fast i dem?
 Er det svært at blive enige i familien om ændringene?
 Hvad betyder det, hvis en ikke synes, at det er en god idé?
5. Eleven og forælderens oplevelse af projektet (30-35 min)
Resten af familien behøver ikke at deltage
Nu vil jeg gerne høre jer [eleven + forælder] fortælle om, hvordan det var at deltage i Step by step projektet.
Elevens spørges først – derefter forælder.
o
Hvordan synes du, at det har været at deltage i projektet?
 Hvorfor?
o
Har du oplevet at projektet har gjort det nemmere at spise og leve sundt? Hvordan?
o
Hvad var det bedste ved at deltage i projektet?
o
Hvad var det værste ved at deltage i projektet?
o
Tænkte du over det med at spise sundt, før projektet startede? Hvis ja: Hvordan? I hvilke situationer?
o
Er du blevet mere interesseret i at spise og leve sundt, end du var før projektet startede? Hvordan?
Hvorfor/hvorfor ikke?
o
Tænkte du på at lade være med at deltage? Hvis ja, hvorfor deltog du alligevel?
Til eleven:
o
Hvordan har det været i klassen?
91






Gik alle op i det?
Hvis ja, Hvorfor tror du, at de gjorde det?
Er det noget, som I har snakket meget om eller kun lidt om?
Hvis snak om konkurrencen, hvor meget betød den?
Hvis den ikke havde været der, hvordan tror du så, det ville have været?
Hvad så nu, snakker I stadig om emnet?
Til eleven:
o
Hvem talte du med om projektet (klassekammerater, din lærer, venner, hele familien, den anden aktive i
familien etc.)?
 Hvad talte I om?/ Hvad fortalte du?
 Støttede de dig i din deltagelse? Hvordan? Hvem støttede mest?
 Eleven: Gjorde dine forældre noget for at du huskede at svare på sms’erne? For at du kunne
opfylde de mål, du havde sat dig? Hvad gjorde de?
 Hvordan har det været, at en anden i familien også har været med i dette projekt?
Til forælder:
o
Hvem talte du med om projektet (XX, hele familien, venner, kollegaer etc.)?
 Hvad talte I om?/ Hvad fortalte du?
 Støttede de dig i din deltagelse? Hvordan? Hvem støttede mest?
 Var der nogen, der hjalp dig til at huske at svare på sms’erne? Var der nogen, der hjalp dig til
at opfylde de mål, du havde sat dig? Hvad gjorde de?
 Hvordan har det været, at XX også har været med i dette projekt?
Til eleven:
Diætistens timer
o
Er der nogle af de ting, som diætisten talte om, som du husker særlig godt?
 Hvilke?
 Hvad var godt og hvad var skidt ved besøget fra diætisten?
 Hvad betyder det, at det er en person som er særligt uddannet inden for kost og ernæring som
fortæller om det frem for andre?
SMS dagbog
o
Hvad er dit indtryk af sms-rapporteringen?
 Hvorfor?
 Hvordan var det at modtage sms’er hver aften?
 Her ser du den besked, du modtog hver aften (vises på interviewers mobiltelefon (+ evt. stykke
papir):
 Hvad tænkte du, da du modtog den?
 Følte du dig kontrolleret eller hjulpet?
 Når du besvarede sms’erne, hvorfor gjorde du det/for hvis skyld gjorde du det?
 Hvad gjorde du for at huske at svare?
 Svarede du på alle sms’er – uafhængigt af, hvad du lavede eller hvor du var?
Elev: Fik du hjælp fra dine forældre?
Forælder: Fik du hjælp fra nogen? Hvem?
Begge: På hvilken måde? (fx påmindelser etc.)
92



Skrev du, hvor meget frugt, grønt eller søde drikke, du havde spist eller drukket eller gættede du
engang imellem?
Gjorde du noget for at huske, hvor meget frugt, grønt og søde drikke du spiste og drak?
Gjorde du noget for, at der var mere frugt og grønt og mindre søde drikke i huset?
 Hvad gjorde du?
Elev:
o
Hvordan satte du dine mål for frugt, grønt og søde drikke? (var det fx en kollektiv ting i klassen, hvor
læreren hjalp?)
 Huskede du på kostrådet med ”6 om dagen”, når du satte dine mål?
 Snakkede du med dine klassekammerater om deres mål, før du satte dine?
Forælder:
o
Hvordan satte du dine mål for frugt, grønt og søde drikke?
 Huskede du på kostrådet med ”6 om dagen”, når du satte dine mål?
 Snakkede du med XX om hans/hendes mål, før du satte dine egne?
Begge:
o
o
Var der nogle ting, der gjorde det nemmere for dig at opnå dine mål?
Ændrede du dit mål, når det var muligt?
 Hvis ja, hvordan kan det være?
 Hvis ja, Stillede du større eller mindre krav til dig selv?
 Kan du huske hvad dine mål for frugt, grønt og søde drikke var? (spørg ind til, hvad et evt.
bekræftende svar går på: spørgeskema 1, 2 el. sms-delen)
 Hvis ja, hvad var målene?
 Nåede du dine mål?
 Hvis ja, hvordan? (var målene fx ikke særlig ambitiøse etc....?)
 Hvis nej, hvorfor ikke?
o Hvad syntes du om perioderne med sms-rapportering? Varede det for kort eller for lang tid?
 Oplevede du nogen problemer med sms-rapporteringen? Hvis ja:
 Hvad skete der?
 Hvordan havde du det med disse problemer? Hvad betød det?
 Ville det være en god idé at lave en periode med sms’er igen for at hjælpe dig til at spise frugt
og grønt?
 Hvorfor god/dårlig idé?
 Hvis god: Tror du også, at det ville være en god idé, hvis der ikke var nogen biografbilletter
eller konkurrence om billetter til Tivoli Friheden?
 Tror du, at der er andre måder, der ville virke bedre – eller lige så godt – som sms’er for at
hjælpe dig til at spise (mere) frugt og grønt? ’
o
Referere til projektet og SMS dagbogen: Hvad kunne ellers hjælpe dig til at spise mere frugt og grønt?
6. Afrunding (5 min)
Nu er jeg ved at have alle de informationer, som jeg havde tænkt mig – og tiden er også ved at være
brugt. Jeg vil gerne sige tak for, at vi måtte komme hjem til jer og gennemføre dette interview. Tak
fordi I ville deltage.
93
Appendix B. Transcription guide (in Danish)
Guide til transkription ved eksempler
Generelt: Lydfilen transskriberes ordret. Man pynter ikke på sproget, selv om det hverken
er velformuleret, flydende eller sprogligt korrekt.
I løbet af interviewene kan I løbende notere, hvem der taler og markere overgange. Det
vil være en stor hjælp, når I efterfølgende skal transskribere.
I:
Interviewer
M:
Interviewperson 1 (manden)
K:
Interviewperson 2 (kvinden)
Kursiv
Ændring i tonefald
Nej det synes jeg du har lært mig godt nok op til efterhånden
FED, CAPS
NEJ selvfølgelig smider jeg ikke glas i skraldespanden
Højrystet i forhold til
resten af sætningen
: (kolon)
Ne:j, jeg bruger bilen på arbejdet
Ordet trækkes ud
(Fx sfa tøven)
(h)
Kort latter
efterhånden
Nej (h) det synes jeg du har lært mig godt nok op til
(hhh)
Lang latter
…
Kort pause
steder
Ja det gør vi i hvert fald … og så handler vi jo også alle andre
………
Længere pause
94
()
Jeg synes, du taler (utydeligt) nu og da
Parentesernes indhold angiver, hvad der med sandsynlighed er blevet sagt
(...)
Ikke muligt at høre hvad der blev sagt, antal prikker = subjektivt vurderet længde af det
utydelige uddrag
M: Vi køber ikke ret meget økologisk kød, men vi køber altid frilandskød (…) fra et
landbrug i Horsens hvor vi har købt igennem et par år, og det er altså grise der går ude i
det fri med lidt skov og lidt mark (……) de får altså ikke penicillin andet end når de er
syge, og de får ikke vækstfremmere, det må de i øvrigt ikke (.) de er så ikke økologiske
men smagen er god...
[
[
Deltagerne taler samtidig:
Angiver, hvor overlappet starter
M: Altså på et tidspunkt der hentede jeg papirer hjem for at investere i sådan en
vindmølle .. (for at energien kunne blive lidt billig) .. i de her huse vi bor i, der bruger vi
[en masse strøm ..
K:
[dvs. nok nærmere børnene
[[ ]]
Transskribentens beskrivelse af andre udtryk, indtryk eller hændelser i
interviewsituationen
[[M rejser sig op for at lukke døren til det tilstødende værelse]]
Udarbejdet af Alice Grønhøj, Institut for Marketing og Statistik, ASB
Tilpasset efter Silverman, D. (1993). Interpreting qualitative data. Methods for analysing
talk, text and interaction. London: Sage Publications.
95
Appendix C. Instruction leaflet (in Danish) handed out to participants in group 2 and 3
(intervention groups)
96
97
Appendix D. Table (in Danish) handed out to participants to support reporting in units
98
Appendix E. Flyer (in Danish) handed out by dieticians to participants in group 3
99
100
4. RELATIONS BETWEEN RESEARCH PAPERS
The three research papers in this thesis contribute to a broader
understanding of adolescent healthy eating by addressing the development of eating
behaviours, the factors influencing healthy eating and the adolescents’ motivation for
behaviour change.
Research paper 1 aims at answering research question 1: Which roles do
parents and adolescents have in healthy eating socialisation. Based on qualitative
interviews with 38 families, the paper explores adolescents’ and parents’ awareness of
and involvement in healthy eating and investigates how they relate it to their roles in the
healthy eating socialisation taking place within the family. The paper takes an intrafamiliar systemic approach when studying healthy eating socialisation and gives a
thorough empirical account. Adolescents were found to adopt two roles: a direct one
placing demands on parents or a cooperative one helping the parents. Parents initiated
dialogues with family members about healthy eating and felt responsible as role models
often fulfilling the adolescent’s demands and acknowledging their help. The findings
confirm that parents still have the upper hand, when it comes to healthy eating, but with
adolescents as active players confirming the notion of consumer socialisation as
bidirectional processes. The findings are also valuable for identifying and understanding
barriers and facilitators of adolescents’ healthy eating.
Research paper 2 aims at answering research question 2: How does the
social influence from parents and friends compared to personal factors impact
adolescents’ healthy eating? Through statistical analysis of baseline questionnaire data
from both adolescents and parents, the paper tests whether the common belief that
children become increasingly influenced by friends at the expense of parents during
101
adolescence is also true for healthy eating. The paper draws on social cognitive theory
and the focus theory of normative conduct and finds that parents remain the main
influencer, with what they do (descriptive norms) being more important than what they
say (injunctive norms). The study contributes to a more comprehensive understanding of
what influences adolescent healthy eating by comparing the influence of entangled social
spheres (parents and friends) while also controlling for personal factors such as the
adolescent’s self-efficacy and outcome expectations.
Research paper 3 aims at answering research question 3: How can a
feedback intervention based on socio-cognitive theory and using text messaging improve
adolescent healthy eating and why? The paper presents statistical analyses of pre- and
post-intervention questionnaires from adolescents (same sample as in research paper 2) to
investigate the effects of a healthy eating intervention by employing feedback via text
messaging during 11 weeks in order to improve adolescents’ behaviour, self-efficacy and
outcome expectations regarding fruit and vegetable intake. While finding no direct
significant effects of the intervention, the study finds significant effects of the level of
activity in the intervention (reflected in the number of sent text messages by participants)
on intervention outcomes. Participants sending more than half of the possible text
messages significantly increased their fruit and vegetable intake. Participants sending
between 10 and 50% of the possible text messages experienced a significant drop in selfefficacy and those sending less than 10% experienced a significant drop in outcome
expectations. The findings suggest that participants’ active engagement in an intervention
is crucial to its success which has implications for health-promoting interventions.
By taking a social cognitive approach when studying healthy eating, it is
possible to reflect on the development of healthy eating through consumer socialisation
processes (research paper 1), the social influences from parents and friends (research
102
paper 2) and behaviour change via a feedback intervention (research paper 3). The social
cognitive approach has proved valuable since linking the establishment of behaviours and
behaviour change provides a richer understanding of adolescent healthy eating.
103
104
5. FAMILY MEMBERS’ ROLES IN HEALTHY-EATING
SOCIALISATION BASED ON A HEALTHY-EATING
INTERVENTION1
Abstract
Purpose: Healthy-eating socialisation is often described as a bi-directional
process, but there are only few studies on children and parent’s roles in the process. This
paper aims to investigate children and parents’ accounts of awareness and involvement in
healthy eating and how they relate it to their roles in healthy-eating socialisation.
Design/methodology/approach: Semi-structured interviews were conducted
with 38 families three months after a healthy-eating intervention involving dietary advice
and SMS feedback. The interviews were analysed by means of qualitative content
analysis.
Findings: Children and parents identified several causes of awareness and
involvement in healthy eating: new or re-activated health knowledge, visualisation of
amounts, self-regulation and planning. Children adopted two strategies in terms of family
socialisation: a direct strategy placing demands on parents or a cooperative strategy
helping the parents. Parents initiated dialogues with family members about healthy eating
and felt responsible as role models often honouring the children’s demands and help.
1
This chapter was published as: Pedersen, S., Grønhøj, A., & Bech-Larsen, T. (2012). Family
members' roles in healthy-eating socialization based on a healthy-eating intervention. Young
Consumers: Insight and Ideas for Responsible Marketers, 13(3), 208-223.
105
Research limitations/implications: Findings provide a concrete empirical
account of the socialisation process and confirm that parents still have the superior hand,
when it comes to healthy eating, but with children as active players. The authors suggest
future studies to explore the development of influence and awareness of healthy eating
among children and the extent to which children wish to engage in healthy-eating
socialisation.
Originality/value: The study supplements previous research by including
children’s immediate family as a unit of analysis. By taking an intra-familiar systemic
approach to studying family socialisation, future studies can take into account the family
support (or lack hereof), when designing interventions and evaluating the outcomes.
5.1. Introduction
Soaring overweight and obesity levels among children in the western world
are well-described in the literature as a serious challenge for the health of future
generations (Reisch & Gwozdz, 2011) with the risk of bringing the undesirable eating
behaviour into adulthood (Klein-Hessling, Lohaus, & Ball, 2005; Knai, Pomerleau, Lock,
& McKee, 2006; Stice, Shaw, & Marti, 2006). Much previous health research emphasises
the important role of parents when it comes to inducing children towards healthy-eating
habits (Birch & Fisher, 1998; Kremers, Brug, de Vries, & Engels, 2003; Lau, Quadrel, &
Hartman, 1990). Parental influence (Eldridge & Murcott, 2000) or parenting style
(Arredondo et al., 2006) embedded in the family’s interaction and communication
patterns (Nørgaard & Brunsø, 2011; Nørgaard, Brunsø, Christensen, & Mikkelsen, 2007)
and as a part of the overall consumer socialisation (Kelly, Turner, & McKenna, 2006)
seem to be important for transmitting healthy-eating habits to children.
106
School-based interventions are another means to create awareness and
behavioural changes with regard to healthy eating, and in recent years, interventions
aiming at improving children’s dietary patterns have become widely applied. These
interventions directly or indirectly aim at inducing participants to change their behaviour
during a relatively short time span and intervention outcomes are usually measured in
terms of changes in Body Mass Index (Stice et al., 2006), healthy food consumption
levels (Pearson, Atkin, Biddle, & Gorely, 2010; Pearson, Biddle, & Gorely, 2009) or
levels of knowledge about healthy eating (Blanchette & Brug, 2005; Shepherd et al.,
2006).
Previous studies strongly suggest that the family is decisive as gatekeepers
in terms of the barriers and facilitators for children’s healthy eating (Birch & Fisher,
1998). But when it comes to healthy-eating interventions as a socialisation attempt
involving children there is a lack of knowledge about how these interventions affect or is
affected by family interaction. Arguably, a family perspective is important for
understanding how an intervention is perceived by and possibly impacts both the
participating child and its immediate family. Hence, studies of healthy-eating
interventions, especially those targeting children, should explicitly take the family context
into account, both to be able to better assess the healthy-eating socialisation processes in
families and in order to design more effective interventions in the future. The present
paper is a contribution to fill this gap in previous research. Based on our review of the
literature, we know that children and parents influence each other mutually when it comes
to healthy food consumption, but there is a lack of knowledge about the roles children and
parents play in healthy eating socialisation. Hence, the purpose of this paper is to
investigate children’s and parents’ awareness of and involvement in healthy eating after
participation in a healthy-eating intervention targeted at the child. An additional aim is to
107
explore how family members perceive their own roles in the process of healthy-eating
socialisation. Besides contributing to knowledge about family members’ roles in healthyeating socialisation, this research is useful for designing future interventions that take the
support (or lack hereof) in a family setting into account. Notice in this connection that the
fact that we study these processes after an intervention is just characteristics of the
context. In this article, we do not focus on, nor do we report, quantitative effects of the
specific healthy-eating intervention. For such results, we refer to other publications from
the same research project (Bech-Larsen & Grønhøj, submitteed). In this article, we report
an account of the children’s and parents’ views on healthy eating and socialisation based
on qualitative research.
5.2. Family members’ awareness of and involvement in healthy-eating interventions
The World Health Organization defines health as “a state of complete
physical, mental, and social well-being and not merely the absence of disease or
infirmity” (WHO, 2011). Healthy eating is behaviour securing a properly functioning
body through intake of healthy food and drink. In this paper, healthy eating will be more
narrowly defined as complying to the official Danish health recommendations (eight
guidelines recommending eating fish, fibre, fruit and vegetables, drinking water, exercise,
avoid fat and sugar and eating a varied diet) (Ministry of Food, Agriculture and Fisheries,
2011).
Healthy-eating interventions attempt to create more awareness of and
involvement in healthy eating among the participants hopefully changing their eating
behaviours in a healthier direction. There is an extensive body of literature focusing on
interventions for both overweight children (Bauer, de Niet, Timman, & Kordy, 2010) and
108
normal weight children (Stice et al., 2006). Interventions take place in different settings,
but are mainly school based (Anderson et al., 2006; De Bourdeaudhuij, 2011; Kubik,
Lytle, Hannan, Perry, & Story, 2003) or family based (Berry et al., 2004). Different tools
as input for each participant are used, such as diet counselling (Eliakim et al., 2002),
tailored information or feedback (Brug, Campbell, & van Assema, 1999; De
Bourdeaudhuij & Brug, 2000). Studies have also explored issues such as parenting style
(Arredondo et al., 2006), family language (Geer, Tulviste, Mizera, & Tryggvason, 2002)
and knowledge levels (Hursti & Sjödén, 1997) in connection to children’s healthy eating.
In all of these studies, the individual child’s behaviour before, during and after the
intervention is usually in focus, rather than that of the whole family. An exception is the
area of nutritional advice and clinical psychology, where studies have been conducted
targeting both children and parents (Wilfley, Vannucci, & White, 2010). When it comes
to preventive healthy-eating interventions, involvement of parents, if any, is usually
restricted to providing their children with information about the intervention (Stice et al.,
2006). This seems insufficient since children are very much affected by the decisions
taken in the family regarding healthy eating and depend on parental support, for instance,
with respect to availability of healthy options at home. Therefore, it is important to study
family members’ awareness of and involvement in healthy eating in connection with a
child’s participation in a healthy-eating intervention. De Bourdeaudhuij
has studied
family members' perceived influence on introducing healthy food into the family and her
findings provide support for targeting families rather than individuals in nutrition
interventions.
From a social learning point of view, behavioural change interventions
should teach participants a more appropriate behaviour or, if their behaviour already is in
line with dietary recommendations, confirm this behaviour. The learning outcome may
109
include experiences with self-regulation, changed self-efficacy levels, increased
knowledge, etc. and can be of a more or less conscious nature. Cognition, behaviour and
environment (peers, family, social structures) are believed to influence the social learning
process in a reciprocal way (Bandura, 1977, 2004; Luszczynska & Schwarzer, 2005)
which makes it a complex task to determine the causality of possible learning outcomes
and effects on behaviour. However, to enable and support the learning process multicomponent interventions are recommended, making it easier for the participant to
implement the learning outcomes with support in different settings such as school, peers
and family (Birnbaum, Lytle, Story, Perry, & Murray, 2002; Müller, Danielzik, & Pust,
2005; Shepherd et al., 2006). Brug et al. (2005) argue that in order to understand the
effects of healthy-eating interventions, including how learning outcomes perceived by
children and parents can be implemented in the participants’ everyday lives, barriers and
facilitators in the surrounding environment (mainly parents in this case) needs to be
investigated.
Hence, in accordance with a social learning approach, we argue that
children’s healthy eating should not be studied in isolation, but take the child’s immediate
family into account. This systemic approach to intra-familiar communication and
interaction is valuable when viewing a healthy-eating intervention as a socialisation
attempt. However, there is a lack of studies looking into the roles of family members in
relation to interventions aiming at creating awareness of and involvement in healthy
eating. Apart from looking at the possible effects of an intervention (in terms of, e.g.,
behavioural changes), it is valuable in its own right to study the amount and nature of
family interaction that is likely to follow from such an intervention.
110
5.3. Family members’ roles in healthy-eating socialisation
Building on classic socialisation theory, consumer socialisation research
places parents as the main agent in the primary socialisation process, where they both
explicitly and implicitly transmit consumer-related orientations to the child (Ekström,
2006; John, 1999; Ward, 1974). In the secondary socialisation process, peers, school and
media become more influential (Dotson & Hyatt, 2005; John, 1999; Moschis, 1985).
Eating healthily is a part of the child’s consumer socialisation process where family,
school, peers etc. play a long-term role in influencing the child’s eating habits (Chan,
Prendergast, Grønhøj, & Bech-Larsen, 2009; Kelly et al., 2006; Murnane, 2008).
Children’s cognitive abilities are important factors in family interaction and the process of
consumer socialisation. For example, children aged 7-11 are at the ‘analytical stage’,
where information processing abilities are improved, more complex knowledge is taken
in and new perspectives that go beyond their own feelings and motives can be applied
(John, 1999). At the ‘reflective stage’ (age 11-16), the social and cognitive abilities are
further developed, and “attempts to influence parents and friends reflect more social
awareness as adolescents become more strategic, favouring strategies that they think will
be better received than a simple direct approach” (John, 1999). In short, pre-adolescents
and adolescents typically have similar knowledge of and experience with strategies to
influence decision-making processes as those of adults (Easterling, Miller, & Weinberger,
1995; Grolnick, Deci, & Ryan, 1997; Nørgaard et al., 2007).
Scholars agree that nowadays pre-adolescent children play a role in
household decision-making (Dotson & Hyatt, 2005; Easterling et al., 1995; Foxman,
Tansuhaj, & Ekström, 1989; Grønhøj, 2002). However, our knowledge about children’s
influence on family decision making and role in family interaction is still rather
fragmented and superficial. Grønhøj (2006) has explored family interaction in relation to
111
environmentally oriented consumer behaviour and her findings suggest that children
influence their parents’ consumption choices both indirectly and directly contributing to
consumer socialisation of all family members. Nørgaard et al. (2007) investigated
children’s influence on and participation in family decision processes and family conflicts
and conflict resolution (Nørgaard & Brunsø, 2011) related to food buying. Their findings
suggest that children’s active participation determines the influence they gain on food
buying and that they use specific strategies to ease family life. With reference to healthy
eating, Ayadi and Bree (2010) argue that food meal times are ways of socializing family
members in consumption skills related to food and they argue that food learning is a twoway process between children and parents. A study on adolescents’ everyday food
practices suggests that the independency of adolescents in terms of what and when to eat
can create intergenerational conflicts, but also that meal times are a way of maintaining
family relatedness and love . Hence, the active role of children in family interaction
implies that consumer socialisation is not merely a one-way process, but can be ‘reverse’
(Foxman et al., 1989; Grønhøj, 2002; Moschis, 1985) or a collection of ‘bidirectional
interactive processes’ (Kuczynski & Parkin, 2006) where mutual influence and value
exchange take place between parents and children (De Mol & Buysse, 2008; Knafo &
Galansky, 2008). These processes can involve conflicts, but also conflict resolution and
avoidance, communication and influence strategies (Nørgaard & Brunsø, 2011) – which
can be both direct and indirect.
Based on this literature review, we now report out qualitative investigation
of children’s and parents’ accounts of awareness of and involvement in healthy eating
after having participated in a healthy-eating intervention. We also explore how family
members perceive their own roles in the process of healthy-eating socialisation.
112
5.4. Method
5.4.1. Sampling
To investigate family interaction regarding healthy eating, interviews were
carried out with children and their families three months after they had participated in a
healthy-eating intervention. A total of 43 families responded positively to the recruitment
letter which was sent to pupils at 12 schools in the Central Denmark Region that had
participated in the healthy-eating intervention. In 38 of these families, interviews were
carried out in October 2008. The interviews took place in the participants’ home at a time
of their convenience (evening or weekend) with the families. Background characteristics
of participants in the interviews, compared to non-interviewed participants in the
intervention, are shown in table 1.
Table 1
Background information
The interviews had an average duration of 100 minutes and were conducted
by a professional interviewer from a market research agency experienced in interviewing
children and families. Student assistants participated in digitally recording the interviews,
taking pictures, registering the family members’ names, age and occupation and made
observation notes. Subsequently, the interviews were transcribed verbatim.
113
5.4.2. Interview procedures
The interview guide was semi-structured in order to invite interviewees to
answer using their own words and narrative structures (Flick, 2009), thereby taking the
children’s cognitive abilities into account. The interview guide focused on the family’s
habits concerning meals, leisure time, physical activity, grocery-shopping and cooking
with emphasis on responsibility, motivation and initiative. More indirectly, the family
discourse on health was investigated as well as its knowledge of nutrition and health
recommendations. In a fruit and vegetable session (similar to one conducted by a dietician
at school), the family members had to choose what they believed to be 600 grams of fruit
and vegetable and put it on their plate. The plates were then weighed and the difference
calculated. The session served to illustrate six portions of fruits and vegetables which
created a dialogue within the family about their current habits. This also gave a break in
the rather long interview. The last part of the interview guide concerned the family’s
overall evaluation of the intervention. They were asked to evaluate the outcome and relate
it to their family interaction regarding healthy eating.
Children’s attitudes and behaviour can be difficult to study since their
willingness to share thoughts in an interview setting vary with their maturity, roles in the
family and other characteristics (shyness, little involvement etc.). Therefore, the
interviewer was instructed to make sure that all family members were heard and were
given the possibility to answer questions; as a principle, children were asked first and
then the parents. The interviewer was also instructed to interpret answers during the
interview in order to make sure that the statements were correctly understood. Parents
sometimes answered on behalf of the children and here the interviewer asked the child
again. Often parents would also supplement the child’s statements, creating a dialogue.
We acknowledge that the interview can be viewed as an intervention in itself influencing
114
the children’s and parents’ statements, however by following the interview guidelines
described above, we have tried to avoid the interviewer influencing the families.
5.4.3. Background for interviews
As already mentioned, the background for the interviews was a healthyeating intervention, where 242 children (age 10-12, 5th grade) were recruited class-wise
from 12 schools in the Central Denmark Region. Randomly assigned to three
experimental groups, they were subjected to different treatments in a nine-week
intervention period; focus was on intake of fruit, vegetables and (reducing) sweet drinks
(for example soft drinks, ice tea and drinking yoghurts), all referred to here as ‘healthy
eating’. The intervention consisted of education by a dietician and individual reporting
and feedback by SMS (mobile phone Short Message Service). During the intervention all
classes were visited twice (45 minutes) by a dietician from the Centre for Public Health,
Central Denmark Region who, following the official health recommendations (6 portions
of fruit and vegetables per day), educated the children about sugar in food and beverages
and the importance of eating fruit and vegetables. A practical exercise was conducted
where the children were asked to choose what they believed to be 600 grams from five
kilos of mixed fruits and vegetables. The amounts were later on weighed. The exercise
was repeated in the interview session (described below). The content of the education
sessions was coordinated by the first author to ensure that all classes received identical
information.
The second component was individual reporting based on communication
via SMS. Every other week, the participants in two of the three experimental groups had a
daily task of reporting, via SMS, their intake of fruit, vegetables and sweet drinks in units,
115
as instructed in class by research assistants. One unit consisted of 100 grams of fruit or
vegetables or 150 ml of sweet drinks; this was depicted on laminated sheets handed out to
participants. On sending the daily SMS-report, participants received prompt feedback
comparing their report to a self-set weekly goal which could be adjusted between SMSweeks. Parents participating in SMS-reporting were instructed in a letter. Participation
was voluntary and parental consent had to be obtained.
5.4.4. Data analysis
The interviews were analysed by means of qualitative content analysis
(Flick, 2009). Data coding followed the principles of theoretical coding and was based on
concepts from the literature review and the empirical data itself. In order to explore
similarities and differences between children and parents and their accounts of awareness
of and involvement in healthy eating and roles in healthy-eating socialisation, thematic
coding was applied. Since the focus was not on the intervention effects, but on children’s
and parents’ accounts of healthy-eating awareness and involvement and roles in the
process of healthy eating socialisation, no comparisons between groups were conducted.
Also, all interviewed families were included in the analysis, no matter how much they
reported on awareness and roles in healthy-eating socialisation.
Views on health and health behaviour differed a lot; in some families it was
perceived as good health behaviour to eat breakfast containing a lot of wholegrain, lowfat dairy products and no sugar, while in other families the emphasis was on the act of
eating breakfast no matter what it consisted of. Therefore, in the analysis no view on
health was taken at face value. Instead a more holistic approach was taken noting selfcontradictions and observing how the families talked about and handled the fruit and
116
vegetables in the practical exercise. Also, no attempt has been made to rate and compare
the overall health status of the families.
5.5. Results
The empirical findings are depicted in figure 1 which serves as a heuristic
tool (Flick, 2009) providing a structure of associations and links between categories
coded for the analysis. The boxes represent core categories from the literature review:
awareness of and involvement in healthy eating based on a healthy-eating intervention
and family members’ roles in healthy eating socialisation.
The circles represent categories derived from the interviews with the
parents’ categories above the line and the children’s categories below the line. However,
these are not the exact words used by participants, but our categorizing statements and
concepts. The number of children/parents confirming the aspects is illustrated by how
close the circles are to the core categories (for instance, more children confirmed selfregulation than increased will and self-efficacy, but these are still related). Children’s and
parent’s perceptions sometimes overlap, but in order to reveal differences as well as
similarities, children and parents are analysed separately with respect to each category
and its’ subordinated concepts.
117
Figure 1
Associations and links based on the interviews. Parents’ categories are shown above the line, children’s
below the line. HE stands for healthy eating.
5.5.1. Awareness of and involvement in healthy eating based on a healthy-eating intervention
Awareness of and involvement in healthy eating can arise from a number of
occasions over time: Family meal times, advertising, home economics classes, grocery
shopping, diets and so on. According to socialisation theory, parents, friends, school,
media and other family members are the main sources. In this study, all children and
(indirectly) parents had taken part in the healthy eating intervention which naturally led
them to reflect on the outcomes caused by the intervention. It was expected that both
children and parents could have learned something new about healthy eating from the
intervention and tried to influence family members at home. It was also expected that the
intervention could serve as a reminder of tacit knowledge which became activated again.
But the awareness of and involvement in healthy eating could also arise from events
118
before/outside the intervention such as personal experiences with special diets, diseases or
health knowledge from books etc. On a methodological note, the interview session itself
may have triggered a heightened awareness of healthy eating. The following sections
present children’s and parents’ general accounts of awareness and involvement based on
the healthy-eating intervention and the circles in figure 1 represent the categories related
to outcomes by children and parents accordingly.
5.5.1.1. Children
The interviewed children mentioned several accounts of awareness of and
involvement in healthy eating based on their participation in the healthy-eating
intervention. Increased awareness arose from thinking and talking more about healthy
eating. In relation to the intervention they said they had been thinking about the
consequences of unhealthy eating, the benefits from healthy eating and ways to be a bit
healthier by cutting down on sweet drinks and increasing consumption of fruits and
vegetables. For example, a boy said: “Yes, I am thinking about it – that the unhealthy
stuff might taste good, but fruit and vegetables may be better” (boy, 503). Children had
talked with class mates, family and friends as predicted by socialisation theory about
different aspects of healthy eating. One girl explained how she and her friends had spent
quite some time talking about the intervention and personal improvements. Increased and
continuous focus on healthy eating was mentioned as a way of reminding oneself of selfset goals, avoiding situations including unhealthy food or choosing healthier alternatives.
One boy explained that due to a sports injury he could not exercise and therefore he was
more aware of getting his six portions (fruit/vegetables) a day.
119
Knowledge serves as a link between awareness and the possibility of
changing behaviour and thereby more active involvement according to social learning
theory. From the sessions with the dietician, children learnt about sugar levels and the
official Danish health recommendations. One boy said that his newly obtained knowledge
was good for him, since he now knew the sugar levels in some of his favourite products
like chocolate milk, drinking yoghurts and soft drinks. He used the knowledge to refrain
from the products choosing healthier alternatives instead. But knowledge did not only
refer to facts about products; for many of the children knowledge was equivalent to
experience. Through the SMS-based intervention they had experienced setting goals and
living up to them. Successes and failures had made them more knowledgeable about the
difficulties related to eating the recommended quantities of fruit and vegetables or
avoiding sweet drinks. Other children gave the impression that the intervention did not
bring them new knowledge especially because their parents had taught them about
healthy eating by implementing healthy routines at home such as eating fruits before
dinner instead of crackers, only eating sweets once a week etc.
The children heavily underlined that the fruit and vegetable session (both
with the dietician and in the interview) served as bringing awareness to eating healthily.
One boy said: “I was very surprised by (discovering) how much you actually have to eat
to get the 600 grams, and then I thought it was a lot to eat” (boy, 503). Representing the
opposite experience, one girl explained how the weighing session in class surprised her,
since “I thought 600 grams was much more” (girl, 1116). Children largely agreed on the
usefulness of a handed out counting form that clearly visualised one unit of fruit,
vegetables and sweet drink. For those reporting by SMS, it served as support, and several
families had kept the form in sight, to remind them of getting enough fruit and vegetables
and avoiding sweet drinks.
120
When accepting new insights from learning, trying to align behaviour with
the newly acquired knowledge seems reasonable. The children tended to report on
increased self-regulation; substituting foods of a more unhealthy character with fruit,
vegetables or abstaining from sweet drinks, as illustrated by this statement: “On the first
day of the project, we went to McDonald’s, and here I actually thought I did something
by ordering water instead of a soft drink” (boy, 407). Others talked about turning down
soft drinks at family parties or asking for fruits instead of the usual unhealthier snacks
when visiting friends after school. Self-regulation could also be social, as illustrated by a
girl saying how her cosy evenings with friends instead of including the usual crisps and
sweets consisted of a buffet of apples, grapes and homemade buns brought by her friends.
The social aspects of self-regulation almost became a social regulation of group member
behaviour when discussing postponing a birthday party for class mates due to the SMSreporting on sweet drinks. The birthday was held on a Friday based on the reasoning that
the children then had to keep away from sweet drinks for the rest of the weekend. Some
children explained that the sense of self-regulation was strongest during the intervention
and had disappeared afterwards for different reasons such as decreasing interest, no SMSfeedback or because the unhealthier alternatives became too attractive. Some children
reported that their health behaviour was absolutely unaffected by the intervention; already
eating healthily or not wanting to change behaviour were the main reasons. However,
some children actively decided to live healthier thereby stating their active involvement.
This could be a result of the intervention, but with the interviews done three months after
the intervention, it is not possible to tell from the interviews whether there was any longterm impact on behaviour.
The children who actively tried to change health behaviour reported that
they felt an increased will or self-efficacy (understood as belief in their own ability to
121
achieve goals (Bandura, 1977)). Increased will was expressed as making an effort to try
unfamiliar fruits and vegetables; a mother quoted her daughter as saying “Next time we
are having broccoli salad, then I would like to taste it” (mother, 405). Another girl
mentioned that tasting new foods had a positive effect on her intake, but she felt
disappointed with herself for not eating more knowing that it tasted good and was healthy
– the motivation or will was simply not there. Some forced themselves to healthier eating
as illustrated by one boy who recognized that even though he did not like vegetables the
intervention “gave me more will to eat them” (boy, 407). Through goal-setting some felt
an increased belief in achieving self-set goals. Realizing that it was not that hard to eat for
instance six a day made some continue after the intervention, but at the time of the
interviews, most children did not think about the goals anymore.
5.5.1.2. Parents
It was clear that the intervention and the interviews activated pre-obtained
knowledge among the parents renewing the awareness of healthy eating. Some parents
admitted that their previous healthy habits had over time turned in to not so healthy habits
(such as also eating many vegetables for dinner over time was substituted by eating a few
slices of cucumber because that was what the children preferred). The difference between
objectively and subjectively consumed amounts was pointed out by parents – especially
those taking part in the SMS-reporting – as a way of increasing awareness and was
illustrated by one father saying: “Being forced to count concretely and not just what one
subjectively thinks was eaten was a huge eye opener” (father, 901). Therefore, visualising
one unit of fruit/vegetable/sweet drink also made it easier for the parents to handle the
recommended six a day. As with the children, parents valued the handed out counting
form: “When I used to hear about this six a day or six units of fruit and vegetables per
122
day, I thought it was absolutely unrealistic (…) and it was not until receiving this chart
where I… well, I can see the measures and I can see that it is realistic” (mother, 105).
This was also evident during the fruit and vegetable sessions in the interviews. For some
parents the weighing session or the child’s participation in SMS-reporting was a
confirmation of their current lifestyle. These parents clearly stated that the intervention
had not changed their behaviour, but the participation confirmed their assumption that the
child was eating enough fruit and vegetables per day, did not drink too many sweet drinks
etc. This made parents content and proud and for some it justified the occasional intake of
unhealthy foods.
As with the children, parents heavily emphasised the intervention as
creating awareness of eating fruit and vegetables and avoiding sweet drinks. Facts related
to healthy eating were not new to most of the parents, but whereas children mentioned
focused thinking and talking more about healthy eating in everyday settings, parents in
general emphasised planning and/or creating structures that facilitated healthy eating.
Parents mentioned being aware of getting six a day, of spreading the intake of fruit and
vegetables out during a whole day and of how different structuring of the weekdays
influenced their behaviour. One mother said: “It became clear that weekends is the time
where… well, it’s easiest to maintain the structure during the weekdays” (mother, 1006).
Thereby, parents tried to maintain an overview and plan accordingly which can be
ascribed to parental responsibility for providing their offspring with healthy-eating habits.
However, especially the parents taking part in the SMS-reporting also mentioned
awareness about their own eating habits.
Barriers for more involvement in healthy eating mentioned by the parents
were time, prices, motivation and lack of inspiration. Several parents had tried to
overcome some of these barriers by, for instance, buying tasty fruit instead of sweets: “I
123
have to tell myself it is not more expensive than buying a bag of sweets” (mother, 204).
As with the children, it cannot be determined whether behavioural changes are long-term
and necessarily caused by participating in the intervention; however some parents
mentioned that they still avoid sweet drinks or eat bigger amounts of vegetables. As for
the children, the parents’ stated that their sense of self-regulation clearly decreased after
the intervention. One parent explained about her lack of self-control as the reason for
returning to previous routines: “No, it was my fault, that with the Coca-Cola and things
like that, that is because I cannot stay away from it myself” (mother, 107).
More parents than children stated that the intervention did not change their
health behaviour. However, where some children described no changes or no active
health decisions, surprisingly many parents reported on what could be interpreted as
“resistance” to healthier eating. Reasons like convenience, lack of time and motivation
were often mentioned, and one mentioned the need of a wake-up call before changing
habits: “Dad has high cholesterol levels and is a bit overweight and… yes, high blood
pressure and all those symptoms of stress and… the job he has…. but it doesn’t result in
anything. It is 100 times easier to do what you usually do (….) there is no doubt that if
one of us got a really big wake-up call, then….” (mother, 809). Among the interview
persons, two parents had cancer, one had recently had a heart attack (age 44) and one had
had a bypass operation. The two parents diagnosed with cancer said that being diagnosed
for them was a severe warning, and they immediately changed their diet in a healthier
direction. On the contrary, the parent with diabetes said that he knew how to eat
according to the disease: “you have to… but then you learn to bend the rules” (father,
1018). It was clear that perceived consequences of current lifestyle varied a lot among the
interviewed parents and the intervention could not change these perceptions.
124
5.5.2. Family member’s roles in healthy eating socialisation
After these accounts of children’s and parents’ awareness of and
involvement in healthy eating, we will now look into the “black box” of healthy eating
socialisation and family member’s roles in this process. Family members can take on both
active and less active roles, they can be forced into roles and they can try to avoid roles.
As shown in figure 1, parents seem to be mostly involved in the socialisation process
which is in line with common socialisation theory.
5.5.2.1. Children
Based on the children’s awareness of healthy eating, it was clear from the
interviews that children used – to various degrees – this awareness in two ways in the
socialisation process of healthy eating. The children who emphasised awareness of
healthy eating (especially those taking part in the SMS-reporting) took on an active role
and made demands on their parents. The demands mostly centred on provision of food
items and related services with the children acknowledging their secondary role when it
comes to food buying and preparation. One girl concerned about eating enough
vegetables stated her demands quite bluntly: “I said that I wanted more vegetables in my
lunch pack” (girl, 915). Another girl was concerned about the amounts her parents
bought: “Because at the end [of an SMS-week] we could sometimes be running low and
then I had to tell my mum all the time to buy some more apples or vegetables” (girl, 916).
Availability was important to the children, not only when it came to ensuring that there
was enough fruit and vegetables in the house, but also parents making fruit and
vegetables more available and attractive by preparing, peeling, slicing and serving them.
In none of the cases did the children mention asking parents not to buy sweet drinks.
125
The children’s demands included asking for support and help. According to
social learning theory (Bandura, 1977; Luszczynska & Schwarzer, 2005), support from
family and peers is extremely important in order to secure a change of behaviour. Mainly
children participating in the SMS-reporting asked for help in counting units and
remembering to send their text messages. A more general demand was the reassurance
that parents supported and approved their participation in the intervention or in general
their effort of eating healthier. It was clear that children linked performance in the
intervention with family support illustrated by one girl talking about the girl in class
eating most fruit and vegetables: “She is really good … I also think her family is good at
it” (girl, 809). A boy was positively surprised by the fact that his whole family started to
participate in counting units of fruit and vegetables on a sheet of paper at home, when he
started doing the SMS-reports.
Children’s demands on parents had a very immediate impact on family
interaction, since parents had to decide whether to honour the demands or not. A more
indirect role of impacting the socialisation process of healthy eating was by influencing
the family’s eating habits in a more cooperative way. Children actively contributed to the
family’s healthy eating socialisation by putting forward ideas that could facilitate healthy
eating in the family or offer their help. They suggested different ways to prepare fruits
and vegetables, such as cutting different fruits and serving them in a bowl after dinner,
making smoothies for breakfast, and adding more fruit and vegetables to their lunch box.
There were no ideas on how to limit intake of sweet drinks, though. Children also shared
their newly obtained knowledge in their interaction with family. In a family where the
father drank Coca-Cola from a mug every morning instead of coffee, the child could now
criticize the father’s behaviour by providing him with facts about tooth decay as a
possible consequence of his excessive daily intake. Other children used facts from the
126
session with the dietician to remind their families to eat enough fruit and vegetables or to
avoid sweet drinks and explained the consequences. Children’s supportive or cooperative
roles suggest that children are very active players in the healthy eating socialisation of
immediate family members.
5.5.2.2. Parents
As main caretakers and food providers, the parents expressed responsibility
of teaching their children healthy eating habits by being role models. One parent
mentioned preparing the kids for leaving home in such a way that they would know how
to live a healthy life. Some parents also supported the view underlying the intervention
that children could and should learn to take responsibility for their own health by
practising self-regulation and thinking about the consequences of unhealthy eating. Other
parents were sceptical about a project teaching healthy eating to their children, since to
their mind it would take a lot more than just information and/or a short-term intervention.
Political views on who is responsible for the health of individuals (society or the
individual) were also put forward, reflecting an on-going public debate about the role of
the Danish welfare state.
As mentioned in the section on children’s demands on parents, parents
could choose to honour the demands and thereby welcoming the children’s active role.
Many parents did provide children with the desired fruit and vegetables, in larger
quantities and with more variation than usual. As food providers, the parent’s awareness
of healthy eating could regulate a large part of the family’s intake of unhealthy food and
drinks simply by not buying them. It was not necessarily discussed in the family but just
simply carried out as exemplified by one parent wanting to avoid additives in food: “red
127
bangers and cod roe are out (…) and Saturday night sweets are cancelled” (father, 917).
Here the parent displayed a power regulating other family members’ intake of these
foods. But parents also responded positively to children clearly trying to influence the
family’s behaviour in a more healthy direction. One mother said that her boy had bought
a book on smoothies and “then, all of a sudden, we had to buy mangos and other
unfamiliar fruits, didn’t we” (mother, 204). Some parents pointed out barriers for
complying with their children’s wishes: The relatively high prices of fruit and vegetables,
not enough time to prepare it or resistance to dictation from official health
recommendations on how to behave. Some children did not put forward demands about
increased intake of fruit and vegetables during the intervention, but tried to do so during
the interviews. In one case, the parent immediately turned the suggestion down saying:
“Well there you have it: then you have to get an extra lunchbox. Their school bags are
already full... you cannot squeeze more in. It will get squashed, the fruit they are
bringing“ (mother, 204). So here the practical and maybe economic aspects of buying a
bigger lunch box and/or school bag were a barrier for bigger fruit intake.
Parents viewed the socialisation efforts of giving their children healthyeating habits as a continuous, often conflict-ridden struggle. One parent expressed it like
this: “Well, I do hope that maybe sometimes they see that it is not only mum and dad
being stupid when we are saying: now you have to eat this or that” (mother, 811). Parents
heavily underlined the intervention as providing them moral support representing a “third
party” which was harder for the children to argue against, and one mother even said: “I
feel I have been struggling. I know I haven’t been good at making them participate (…)
but it is hard and as a mother you get sick of it and then I feel that the project has made it
a bit easier for me” (mother, 216).
128
Parents’ awareness of and involvement in healthy eating (coming from the
intervention or other events) brought about dialogues between the child and other family
members regarding healthy eating, supporting the parent’s main role as caretakers.
Different types of dialogue were identified in the interviews. One type of dialogue was
about persuasion, illustrated by one parent saying: ”It is about… I think she is easier to
talk into eating fruit and vegetables” (father, 216). Another type of dialogue was about
helping and supporting the child in his or her efforts to change behaviour. Some
dialogues were initiated by the children participating in the intervention making them
report on their new knowledge, recipes, experience, etc. or children asking parents about
health advice. Some parents praised their child when eating healthy, some expressed
normative pressure on the child at mealtimes or when snacking (“we are only saying this
because we love you”), while other parents appealed to the child’s critical sense of what
is right and wrong when raiding the fridge after school.
The dialogues tended to include siblings, regulating their behaviour so that
all children in the family had to abide by the same set of rules. The dialogues gave the
opportunity to discuss conflicting health behaviours such as parents drinking soft drinks
in the evenings while children were not allowed any such drinks. In some cases, the
discussion resulted in group formations where, for instance, the children argued against
their parents, or the mother would argue with the father about being the healthiest or
unhealthiest person in the family. In one case a health behaviour conflict arose during the
interview. One mother was relating her huge efforts in preparing healthy and varied
lunches for her daughter. The daughter cut her off by saying: “But I don’t eat your lunch,
I never did. Well, maybe in the first grade” (girl, 811). The stunned mother, trying to
grasp that for four years her lunches had been dumped, asked her daughter what she then
129
had for lunch. The daughter (very annoyed) answered that she did not want to eat
anything. This statement clearly struck the parents with concern.
The discussions could also be about health beliefs as illustrated by a boy
who cut down on sugary sweet drinks, sweets and crisps. His decision resulted in a
conflict with his mother about him drinking light soft drinks, as his mother preferred that
he got real sugar rather than artificial sweeteners.
The dialogue on healthy eating
revealed different roles in the families. A number of mothers, in particular, defined
themselves as “health promotors” or “family directors” in charge of the family’s health.
Some of them faced resistance from the whole family as illustrated by one mother: “I
have always wanted us to eat fruit. Especially in the mornings (…) and I really would like
us to drink freshly-squeezed juice. But my family definitely do not accept that” (mother,
811). So conflicts also appeared between parents sometimes leading to compromises with
for instance having both raw and cooked vegetables for dinner. In other cases parents
joined forces and pressed the child to improve its’ health behaviour. Some parents were
careful about thrusting too many health messages on-to the child, worried that he/she
would get fed up and refuse to engage in the dialogue. This clearly showed that parents’
took their role as main caretakers and responsible for the children’s healthy eating
socialisation serious, but also that there exist a number of barriers that makes the role hard
to fulfil.
5.6. Discussion
The aim of our study was to study the nature and sources of family
member’s awareness of and involvement in healthy eating and investigate how they
related it to their roles in the process of healthy-eating socialisation.
130
The children’s and parents’ accounts of awareness of and involvement in
healthy eating clearly showed that it can arise from a number of events and to various
degrees: being with friends with healthier habits, reflecting on own health behaviour,
seeing pictures of healthy foods, planning for it etc. As underlined above, it was not this
paper’s purpose to determine the effects of the specific intervention, but give an account
of how children’s and parents’ awareness of and involvement in healthy eating could
crystallize. Among the children, awareness and involvement was characterized by
thinking and taking about healthy eating which could result in self-regulation and noticing
an increased will for healthy eating. Visualisation of amounts of recommended portions
of fruit and vegetables was – according to the children – a successful mean of creating
awareness of healthy eating, and the children converted this into concrete knowledge and
possibly improved behaviour. The children’s accounts of awareness of healthy eating can
be characterized as centred on the child itself; the children mentioned specific events that
made their awareness and involvement go up or down, and it was described as a learning
process. In comparison, the parents’ accounts of healthy eating awareness and
involvement was also characterised by self-regulation and again the visualisation was
effective in providing concrete knowledge of recommended portion sizes. More parents
than children also mentioned that awareness rose by re-activating pre-obtained health
knowledge through different means. This is natural, since parents (presumably) have gone
through a long consumer socialisation process and thereby have been exposed to health
knowledge and messages many more times than their children. When children described
that their awareness development affected themselves, parents described how rising levels
of awareness made them think about planning and structuring healthy eating for the
whole family and thereby regulating the family’s health behaviour as well. Barriers to
healthy eating seemed bigger, when awareness was low, and this also had consequences
131
for the family. Looking at the concrete intervention, including the immediate family in
such an intervention implicates for future intervention designers that it can affect
awareness and involvement levels in a positive direction; both among the directly affected
children, but also among the parents – at least in a short-term perspective as a catalyst for
healthy eating. This suggests that the systemic intra-familiar approach based on social
learning theory is valuable when interested in knowing more about the possible broader
impact of an intervention. However, it is one of this study’s limitations that we are not
able to determine measureable effects of the intervention on awareness and involvement
levels, and therefore we suggest that future studies to do so.
The accounts of children’s and parents’ roles in the process of healthyeating socialisation described different strategies used by the two generations. In the
interaction with parents, children displayed elements of what John called the ‘reflective
stage’ (John, 1999), where attempts were made to influence parents both directly and
indirectly and to different degrees depending on the children’s level of awareness and
involvement in healthy eating. The direct attempts were demands on parents to provide
specific fruit and vegetables in satisfying amounts when the children wanted it. More
indirect attempts of influence were also made, such as suggesting eating healthy foods or
avoiding unhealthy, helping and contributing with knowledge and ideas for the benefit of
the family. These findings are similar to those found by Nørgaard et al. (2011) regarding
food buying. Our results confirmed John’s (1999) description of children at the reflective
stage favouring strategies that they thought would go down better with parents. Our study
also confirmed that children are more likely to gain influence when they participate
actively and help ease family life (Nørgaard & Brunsø, 2011). But whereas Nørgaard and
Brunsø (2011) focused on food buying, both of unhealthy and healthy food, our focus on
healthy consumption leaves little room for children’s strategic considerations about being
132
allowed to eat chocolate bars as well if they are actively involved in the family’s healthy
eating. This we would consider a result of successful socialisation of the child.
Parents did in fact appreciate children’s participation in family health,
valuing the possibility of having a different dialogue with the child and the remaining
family about healthy eating. Sensing a moral support from the intervention (and from
other third parties like schools, media etc.), parents felt confident discussing conflicting
health behaviours and food preferences with the child and the rest of the family, as also
found in Nørgaard et al. (2011). Sometimes this confidence resulted in displaying power
by regulating the whole family’s intake of certain foods which can be considered as
natural having the role as caretaker. Fulfilling children’s demands, parents displayed their
responsibility as caregivers contributing positively to the overall consumer socialisation
of the child (Ekström, 2006), and being a role models for the child. However, parents also
displayed resistance to more involvement in healthy eating. Barriers like time, money and
convenience were given as reasons, and it was surprising to hear parents argue against
their children’s wishes of eating healthier. The barriers could of course be very real in the
particular families, and it displayed that the widespread health discourse in Western
societies, where health as argument seem to overrule everything else (Turner, 2010), is
not necessarily incorporated in all families. It also illustrated that parents have the main
executive power in relation to the socialisation process which simply confirms classical
socialisation theory.
In order to study the socialisation process and family interaction, there are
certain limitations to the method used here. Making interviewees reflect on their own
health and family interaction during an interview with (in most cases) the whole family,
can be problematic, since the “weaker” family members can hold back information.
Observations during interviews can be of too short a duration to capture interaction and
133
they tend to take place with interviewees knowing they are on display. Relying
exclusively on interviewees’ self-assessments can also be problematic since they might
not be able to report on family interaction without actually interacting. Therefore, we
suggest future studies to use additional methods to supplement the “capturing” of family
interaction. Grønhøj and Bech-Larsen suggested the use of vignettes in the study of
family consumption processes (Grønhøj & Bech-Larsen, 2010) by inviting family
members to interact during the interview. This could also be explored in more detail when
it comes to healthy-eating socialisation.
With the children influencing the socialisation process, our study has given
an empirical account with concrete details of the socialisation process which we also
consider as bi-directional (Kuczynski & Parkin, 2006). However, our study showed that
parents still have most power when it comes to healthy eating, no matter how involved
and aware the pre-adolescent children are. We suggest future studies to be longitudinal in
order to explore the development of influence and awareness of healthy eating and
investigate the interesting question of whether children want to engage more in the
process of healthy-eating socialisation and thereby influencing their own health behaviour
more, or whether they prefer leaving it to their role models. This could be conducted with
children at different ages, since we believe (in accordance with John’s description of
children’s cognitive abilities (1999)) that matureness and age are important factors, when
it comes to children’s possibility to gain more power in the socialisation process of
healthy eating.
134
Acknowledgements
The authors would like to thank student assistants Simon Rune Jørgensen,
Pernille Sabroe, Sanne Chrestensen and Helene Olesen for their work in relation to the
data collection for this study. Special thanks to Professor John Thøgersen for valuable
comments on earlier drafts. Also thanks to Birgitte Steffensen for proofreading and to the
Danish Ministry of Science, Technology and innovation for financially supporting the
research project “Step by step changes of children’s preferences towards healthier food”,
grant no. 09/061357.
135
References
Anderson, A. S., Porteous, L. E. G., Foster, E., Higgins, C., Stead, M., Hetherington, M., .
. . Adamson, A. J. (2006). The impact of a school-based nutrition education
intervention on dietary intake and cognitive and attitudinal variables relating to
fruits and vegetables. Public Health Nutrition, 8(6), 650-656.
Arredondo, Elva M., Elder, John P., Ayala, Guadalupe X., Campbell, Nadia, Baquero,
Barbara, & Duerksen, Susan. (2006). Is parenting style related to children's
healthy eating and physical activity in Latino families? Health Education
Research, 21(6), 862-871. doi: 10.1093/her/cyl110
Ayadi, Kafia, & Bree, Joël (2010). An ethnography of the transfer of food learning within
the family. Young Consumers: Insight and Ideas for Responsible Marketers,
11(1), 67-76.
Bandura, Albert. (1977). Social learning theory (1st ed.). Englewood Cliffs, NJ: PrenticeHall.
Bandura, Albert. (2004). Health promotion by social cognitive means. Health Education
& Behavior 31(2), 143-164.
Bauer, S, de Niet, J, Timman, R, & Kordy, H. (2010). Enhancement of care through selfmonitoring and tailored feedback via text messaging and their use in the treatment
of childhood overweight. Patient Education and Counseling, 79, 313-319.
Bech-Larsen, Tino, & Grønhøj, Alice. (submitteed). Promoting healthy eating to children:
Evaluating a text message (SMS) feedback intervention.
Berry, Diane, Sheehan, Rebecca, Heschel, Rhonda, Knafl, Kathleen, Melkus, Gail, &
Grey, Margaret. (2004). Family-based interventions for childhood obesity: A
review. Journal of Family Nursing, 10(4), 429-449. doi:
10.1177/1074840704269848
Birch, L., & Fisher, J. (1998). Development of eating behaviors among children and
adolescents. Pediatrics, 101(3), 539-549.
Birnbaum, A., Lytle, L., Story, M., Perry, C., & Murray, D. (2002). Are differences in
exposure to a multicomponent school-based intervention associated with varying
dietary outcomes in adolescents? Health Education & Behavior 29(4), 427-443.
Blanchette, L., & Brug, Johannes. (2005). Determinants of fruit and vegetable
consumption among 6–12-year-old children and effective interventions to increase
consumption. Journal of Human Nutrition & Dietetics, 18(6), 431-443.
136
Brug, Johannes, Campbell, Marci, & van Assema, Patricia. (1999). The application and
impact of computer-generated personalized nutrition education: A review of the
literature. Patient Education and Counseling, 36(2), 145-156.
Brug, Johannes, Oenema, A., & Ferreira, I. . (2005). Theory, evidence and intervention
mapping to improve behavior nutrition and physical activity interventions.
International Journal of Behavioral Nutrition and Physical Activity, 2(2), 1-7.
Chan, Kara, Prendergast, Gerard, Grønhøj, Alice, & Bech-Larsen, Tino. (2009).
Communicating healthy eating to adolescents. Journal of Consumer Marketing,
26(1), 6-14. doi: 10.1108/07363760910927000
De Bourdeaudhuij, Ilse. (2011). School-based interventions promoting both physical
activity and healthy eating in Europe: A systematic review within the HOPE
project. Obesity Reviews, 12(3), 205-216.
De Bourdeaudhuij, Ilse, & Brug, Johannes. (2000). Tailoring dietary feedback to reduce
fat intake: An intervention at the family level. Health Education Research, 15(4),
449-462. doi: 10.1093/her/15.4.449
De Mol, Jan, & Buysse, Ann. (2008). The phenomenology of children's influence on
parents. Journal of Family Therapy, 30(2), 163-193. doi: 10.1111/j.14676427.2008.00424.x
Dotson, Michael J., & Hyatt, Eva M. (2005). Major influence factors in children's
consumer socialization. Journal of Consumer Marketing, 22(1), 35-42.
Easterling, Debbie, Miller, Shirley, & Weinberger, Nanci. (1995). Environmental
consumerism: A process of children's socialization and families' resocialization.
Psychology and Marketing, 12, 531-550.
Ekström, Karin M. . (2006). Consumer socialization revisited. In R. Belk (Ed.), Research
in Consumer Behavior (pp. 71-98): Emerald Group Publishing Limited.
Eldridge, Jane, & Murcott, Anne. (2000). Adolescents’ dietary habits and attitudes:
Unpacking the ‘problem of (parental) influence. Health, 4(1), 25-49. doi:
10.1177/136345930000400102
Eliakim, Alon, Kaven, Galit, Berger, Isaac, Friedland, Orit, Wolach, Baruch, & Nemet,
Dan. (2002). The effect of a combined intervention on body mass index and
fitness in obese children and adolescents - a clinical experience. European Journal
of Pediatrics, 161(8), 449-454. doi: 10.1007/s00431-002-0980-2
Flick, U. (2009). An introduction to qualitative research (4th ed.). London: Sage
Publications.
137
Foxman, Ellen R., Tansuhaj, Patriya S., & Ekström, Karin M. (1989). Family members'
perceptions of adolescents' influence in family decision making. Journal of
Consumer Research, 15(4), 482-491.
Geer, Boel De, Tulviste, Tiia, Mizera, Luule, & Tryggvason, Marja-Terttu. (2002).
Socialization in communication: Pragmatic socialization during dinnertime in
Estonian, Finnish and Swedish families. Journal of Pragmatics, 34(12), 17571786. doi: 10.1016/s0378-2166(01)00059-5
Grolnick, W., Deci, E., & Ryan, R. (1997). Internalization within the family: The selfdetermination theory perspective. In J. Grusec & L. Kuczynski (Eds.), Parenting
and children’s internalization of values. A handbook of contemporary theory (pp.
135-161). New York: John Wiley & Sons.
Grønhøj, Alice. (2002). Miljøvenlig adfærd i familien: Et studie af familiemedlemmernes
involvering og sociale interaktion [Pro-environmental behaviour in the family: A
study of family members' involvement and social interaction]. Århus: Institut for
Markedsøkonomi.
Grønhøj, Alice. (2006). Communication about consumption: A family process
perspective on “green” consumer practices. Journal of Consumer Behaviour, 5(6),
491-503.
Grønhøj, Alice, & Bech-Larsen, Tino. (2010). Using vignettes to study family
consumption processes. Psychology & Marketing, 27(5), 445-464.
Hursti, U., & Sjödén, P. O. (1997). Changing food habits in children and adolescents.
Experiences from intervention studies. Scandinavian Journal of
Nutrition/Näringsforskning 41, 102-110.
John, Deborah Roedder. (1999). Consumer socialization of children: A retrospective look
at twenty-five years of research. Journal of Consumer Research, 26(3), 183-213.
Kelly, James, Turner, Jason J., & McKenna, Kirsty (2006). What parents think: Children
and healthy eating. British Food Journal, 108(5), 413-423.
Klein-Hessling, J., Lohaus, A., & Ball, J. (2005). Psychological predictors of healthrelated behaviour in children. Psychology, Health & Medicine, 41, 31-43.
Knafo, Ariel, & Galansky, Neta. (2008). The influence of children on their parents’
values. Social and Personality Psychology Compass, 2(3), 1143-1161. doi:
10.1111/j.1751-9004.2008.00097.x
Knai, C., Pomerleau, J., Lock, K., & McKee, M. (2006). Getting children to eat more fruit
and vegetables: A systematic review. Preventive Medicine, 42, 85-95.
138
Kremers, Stef P. J., Brug, Johannes, de Vries, Hein, & Engels, Rutger C. M. E. (2003).
Parenting style and adolescent fruit consumption. Appetite, 41(1), 43-50.
Kubik, Martha Y., Lytle, Leslie A., Hannan, Peter J., Perry, Cheryl L., & Story, Mary.
(2003). The association of the school food environment with dietary behaviors of
young adolescents. Am J Public Health, 93(7), 1168-1173. doi:
10.2105/ajph.93.7.1168
Kuczynski, L., & Parkin, M. C. (2006). Agency and bidirectionality in socialization. In J.
E. Grusec & P. D. Hastings (Eds.), Handbook of socialization: Theory and
research (pp. 259-283). New York: The Guilford Press.
Lau, Richard R., Quadrel, Marilyn Jacobs, & Hartman, Karen A. (1990). Development
and change of young adults' preventive health beliefs and behavior: Influence
from parents and peers. Journal of Health and Social Behavior, 31(3), 240-259.
Luszczynska, A., & Schwarzer, R. (2005). Social cognitive theory. In M. Conner & P.
Norman (Eds.), Predicting health behaviour: Research and practice with social
cognition models (2nd ed., pp. 127–169). Berkshire: Open University Press.
Moschis, G. (1985). The role of family communication in consumer socialization of
children and adolescents. Journal of Consumer Research, 11(4), 898-913.
Murnane, Jennifer Aden (2008). Resocializing adults for their new role as consumercitizens. Journal of Family and Consumer Sciences, 100(4), 10-16.
Müller, Manfred J., Danielzik, Sandra, & Pust, Svenja. (2005). School- and family-based
interventions to prevent overweight in children. Proceedings of the Nutrition
Society, 64, 249-254.
Nørgaard, Maria Kümpel, & Brunsø, Karen. (2011). Family conflicts and conflict
resolution regarding food choices. Journal of Consumer Behaviour, 10, 141-151.
doi: 10.1002/cb.361
Nørgaard, Maria Kümpel, Brunsø, Karen, Christensen, Pia Haudrup, & Mikkelsen,
Miguel Romero. (2007). Children’s influence on and participation in the family
decision process during food buying. Young Consumers, 8(3), 197-216.
Pearson, Natalie, Atkin, Andrew J , Biddle, Stuart JH , & Gorely, Trish (2010). A familybased intervention to increase fruit and vegetable consumption in adolescents: A
pilot study. Public Health Nutrition 13 (6), 876-885.
Pearson, Natalie, Biddle, Stuart JH , & Gorely, Trish (2009). Family correlates of fruit
and vegetable consumption in children and adolescents: A systematic review.
Public Health Nutrition, 12(2), 267-283.
139
Reisch, L., & Gwozdz, W. (2011). Chubby cheeks and climate change: Childhood obesity
as a sustainable development issue. International Journal of Consumer Studies,
35(1), 3-9.
Shepherd, J, Harden, A, Rees, R, Brunton, G, Garcia, J, Oliver, S, & Oakley, A (2006).
Young people and healthy eating: A systematic review of research on barriers and
facilitators. Health Education Research, 21(2), 239-257.
Stice, Eric, Shaw, Heather, & Marti, C. Nathan. (2006). A meta-analytic review of
obesity prevention programs for children and adolescents: The skinny on
interventions that work. Psychological Bulletin, 132(5), 667-692.
Turner, R. (2010). Discourses of consumption in US-American culture. Sustainability, 2,
2279-2301.
Ward, Scott. (1974). Consumer socialization. Journal of Consumer Research, 1(2), 1-17.
WHO. (2011). What is the definition of health? Retrieved January 27, 2015, from
http://www.who.int/suggestions/faq/en/
Wilfley, Denise, Vannucci, Anna, & White, Emily. (2010). Early intervention of eatingand weight-related problems. Journal of Clinical Psychology in Medical Settings,
17(4), 285-300. doi: 10.1007/s10880-010-9209-0
140
6. FOLLOWING FAMILY OR FRIENDS. SOCIAL NORMS IN
ADOLESCENT HEALTHY EATING1
Abstract
It is commonly believed that during adolescence children become
increasingly influenced by peers at the expense of parents. To test the strength of this
tendency with regards to healthy eating (fruit and vegetable intake), a survey was
completed by 757 adolescent-parent dyads. Our theoretical framework builds on social
cognitive theory and the focus theory of normative conduct, and data are analysed by
means of confirmatory factor analysis and structural equation modelling. The study
reveals that when it comes to adolescents’ fruit and vegetable intake, parents remain the
main influencer, with what they do (descriptive norms) being more important than what
they say (injunctive norms). The study contributes to a more comprehensive
understanding of what influences adolescent healthy eating, including the social influence
of parents and friends, while also taking adolescent self-efficacy and outcome
expectations into account. No previous studies have included all these factors in the same
analysis. The study has a number of important implications: (1) healthy eating
interventions should aim at strengthening self-efficacy and positive outcome expectations
among adolescents, (2) the family context should be included when implementing healthy
eating interventions and (3) parents’ awareness of their influence on children’s healthy
eating should be reinforced.
1
This chapter was published as: Pedersen, S., Grønhøj, A., & Thøgersen, J. (2015). Following
family or friends. Social norms in adolescent healthy eating. Appetite, 86, 54-60.
141
6.1. Introduction
Eating practices established in childhood are often carried into adulthood
(Lake et al., 2004). Hence, it is important to establish healthy eating practices early in
childhood and to support them during adolescence (WHO, 2000). Especially, eating
sufficient quantities of fruit and vegetables contributes to the prevention of chronic
diseases and the avoidance of obesity in general (WHO, 2003). Children most often eat in
a social context. They are strongly influenced by parents’ attitudes and behaviours, and as
primary socialisation agents (John, 1999) parents are gatekeepers of their children’s
healthy eating (Birch & Fisher, 1998). As the child grows older, secondary socialisation
agents such as friends, school and media influence behaviour as well (Chan, Prendergast,
Grønhøj, & Bech-Larsen, 2010). Parental influence is believed to decline or at least
change as the child moves into adolescence (Gitelson & McDermott, 2006).
Among the many routes to healthy eating, special attention has been
devoted to increasing the intake of fruit and vegetables – and hopefully replacing
unhealthy food. Although we acknowledge that the latter cannot be taken for granted, and
that reducing unhealthy eating is an important topic in its own right, this study’s point of
departure is the fact that most adolescents do not eat the recommended amount of fruit
and vegetables (Rasmussen et al., 2006; WHO, 2004) and there is a need for a better
understanding of why. Specifically, there is a lack of research on the relative importance
of adolescents’ personal motivation and the social influence of parents and friends on
adolescents’ healthy eating. Therefore, the purpose of this study is to determine the social
influence of parents and friends on adolescents’ healthy eating, specifically fruit and
vegetable intake, while also taking into account adolescents’ personal motivation to eat
fruit and vegetables. A range of motives for food intake has been identified by previous
research (e.g. Herman, Roth, & Polivy, 2003). Bandura’s (1986) Social Cognitive Theory
142
(SCT) is a popular framework for studying people’s motivation to change behaviour (in
our case, increasing fruit and vegetable intake). Many previous studies have confirmed
the importance of the key motivation constructs proposed by SCT, namely self-efficacy
and outcome expectations, for healthy eating (e.g. Fitzgerald, Heary, Kelly, Nixon, &
Shevlin, 2013; Geller & Dzewaltowski, 2010). Self-efficacy is the belief “that one has the
power to produce desired changes by one’s actions” (Bandura, 2004, p. 144). Relevant
outcome expectations regarding an anticipated behaviour are classified into three types:
physical, social and self-evaluative (Bandura, 1977). Further, SCT suggests that a
person’s behaviour is not the product of personal motivation alone, but also learned
through observing the behaviour of others and influenced by perceived social pressure.
The individual’s self-efficacy, outcome expectations and social influence (i.e., perceived
social norms) together lead to behavioural goals or intentions which together with
facilitating and/or impeding contextual factors lead to behaviour.
A common definition of social norms is “rules and standards that are
understood by members of a group and that guide and/or constrain social behaviour
without the force of laws” (Cialdini & Trost, 1998, p. 152). Cialdini and colleagues
distinguish between descriptive and injunctive norms (1991; 1990). Descriptive norms
refer to what is commonly done, whereas injunctive norms refer to commonly held
perceptions of do’s and don’ts. In the context of SCT, it is not so much other people’s
objective behaviour or expectations as the individual’s subjective perception of these
realities that are assumed to influence behaviour (Thøgersen, 2008).
Healthy eating (Fitzgerald et al., 2013) and specifically fruit and vegetable
consumption among adolescents have been found to increase with self-efficacy
(Rasmussen et al., 2006; Young, Fors, Fasha, & Hayes, 2004) and with positive outcome
expectations (Resnicow et al., 1997). As regards social influence on children’s healthy
143
eating, the importance of parents is widely recognised (Lau, Quadrel, & Hartman, 1990)
and parental influence in childhood seems to have long-term effects (Bauer, Laska,
Fulkerson, & Neumark-Sztainer, 2011; Lake et al., 2004). Not surprisingly, given
children’s daily exposure to parents’ attitudes and behaviour, parental intake (Rasmussen
et al., 2006) and adolescents’ perception of parents’ intake of fruit and vegetables
(Kristjánsdóttir, De Bourdeaudhuij, Klepp, & Thorsdóttir, 2009; Young et al., 2004) are
also positively correlated with adolescents’ intake. Adolescents and their parents usually
live together and share the fruit and vegetables that are available in the home and also a
more general context and culture of eating, preparing and planning food intake. SCT
refers to this shared context which may account for some of the similarity in behaviour
between adolescents and their parents, as (facilitating or impeding) socio-structural
factors (Bandura, 1986).
Previous research has also found correlations between adolescents’ and their
friends’ eating behaviour (Bruening et al., 2012) suggesting that friends influence each
other (Ball, Jeffery, Abbott, McNaughton, & Crawford, 2010; Salvy, de la Haye, Bowker,
& Hermans, 2012) and/or conform to common norms (Stead, McDermott, MacKintosh,
& Adamson, 2011). It has also been found that friends influence healthy eating negatively
(Fitzgerald et al., 2013) by sometimes encouraging adolescents to consume unhealthy
foods (Croll, Neumark-Sztainer, & Story, 2001). Others have found that friends restrict
each other’s intake of unhealthy foods (Howland, Hunger, & Mann, 2012) and that
friends’ negative influence can be counteracted by the adolescent’s impression
management concerns (Salvy et al., 2012). Of course, it is not always clear from the
literature whether friends actually influence each other or whether they become friends
based on behavioural similarities (see for instance Bruening et al., 2012).
144
In this paper, the importance for adolescents’ fruit and vegetable intake of
both parents’ and friends’ descriptive and injunctive norms as well as the adolescent’s
own self-efficacy and outcome expectations is investigated. On the basis of the literature,
we expect that all of these variables will influence adolescents’ intake of fruit and
vegetables and that family norms will influence adolescents’ healthy eating more than
their own self-efficacy and outcome expectations will. Specifically, we hypothesize that:
Hypothesis 1a. Adolescents’ intake of fruit and vegetables depends on their
self-efficacy and outcome expectations as well as on the dominant family norms as
reflected in parental behaviour.
Hypothesis 1b. Adolescents’ intake of fruit and vegetables depends more on
the dominant family norms than on their own self-efficacy and outcome expectations.
According to SCT and empirical research (Baker, Whisman, & Brownell,
2000; McClain, Chappuis, Nguyen-Rodriguez, Yaroch, & Spruijt-Metz, 2009) it is the
perception of others’ behaviour more than others’ actual behaviour that influences a
person’s behaviour. Hence, we expect that adolescents’ behaviour will be more strongly
related to their subjective perception of their parents’ behaviour than to their parents’
actual behaviour, and even more so when parent’s actual behaviour is measured
imperfectly by parental self-report.
Hypothesis 2. Adolescents’ intake of fruit and vegetables depends more on
how they perceive their parents’ behaviour than on the parents’ actual behaviour, as
measured by parental self-report.
Since adolescents consume most meals in the family, parents are expected
to be more influential than friends are when it comes to adolescents’ eating. Hence, we
hypothesize that:
145
Hypothesis 3. Adolescents’ intake of fruit and vegetables is influenced more
by the dominant norms in their own family than by the norms that they perceive as
dominant among their friends.
A recent study among adolescents (16 to 19 years old) found that
descriptive norms, but not injunctive norms of peers in school were associated with their
own fruit and vegetable intake (Lally, Bartle, & Wardle, 2011). Hence, we hypothesize
that descriptive norms influence adolescents’ healthy eating more than injunctive norms.
Hypothesis 4. Adolescents’ intake of fruit and vegetables depends more on
what their parents and peers do (i.e., descriptive norms) than on what they say (i.e.,
injunctive norms).
6.2. Methods
6.2.1. Participants and procedure
A sample of 1321 adolescents and 795 parents was recruited from 17
schools in the Central Denmark Region in September 2010.2 The sample contained a total
of 757 adolescent-parent dyads, which were identified by a unique ID number. In the
adolescent-parent dyads sample, there were 347 boys/410 girls and 634 mothers /113
fathers (see Table 1). Hence, girls are slightly and mothers heavily overrepresented in the
sample. Participation was voluntary and no compensation was offered.
2
The present study was carried out as part of the Step-by-Step Project (grant number 09/061357
from the Danish Ministry of Science, Technology and Innovation) which also contained an
intervention study aiming at increasing fruit and vegetable intake among school children
(Pedersen, Grønhøj, Bech-Larsen & Thøgersen, manuscript in preparation).
146
Table 1
Background information
Number of participants
Gender and class level
No. of male/female
Parent
757
Adolescent
757
113/634
(10 missing)
347/410
No. of 5th / 9th graders
No. of boys/girls in 5th grade
No. of boys/girls in 9th grade
Mean age
42.7
Standard deviation
5th graders
9th graders
Parents’ educational level (%)
7th -10th grade
High school or similar
Vocational education
Short further education (up to two years)
College degree completed (2-4 years)
Graduate school (Masters, Doctorate or equivalent,
4 years or more)
Monthly household income before tax
Less than 8,000 DKK/1,441$
8,000-14,999 DKK /1,441-2,702$
15,000-29,999 DKK/2,702-5,404$
30,000-49,999 DKK/5,404-9,007$
50,000-69,999 DKK/9,007-12,609$
70,000 DKK or more/12,609$ or more
5.241
453/304
209/244
138/166
12.5
1.978
10.89
14.83
8.9
7.3
28.0
11.0
34.8
10.0
0.7
3.1
13.6
28.8
35.7
18.0
The questionnaire was thoroughly pre-tested. A school gave access to four
children (age 11) who filled in the questionnaire and afterwards wording and scales were
discussed with the first author. Following adjustments, four new pupils from the same
school went through the same pre-test. A third pre-test was conducted with 30 pupils (age
10-16) who filled in the questionnaire; subsequently frequencies and scale reliability were
checked. The adaptations and translations of existing scales from English into Danish
included back translations and group discussions (Brislin, 1970). For the final data
collection, adolescents completed a questionnaire at school under supervision of a
research assistant while parents completed it at home.
147
6.2.2. Measures
Adolescents and parents answered the same questions about behaviours,
self-efficacy and outcome expectations related to fruit and vegetable consumption. The
adolescents were also asked about their perception of parents’ and friends’ behaviour and
attitudes in relation to fruit and vegetable consumption (see Table 2). Except where other
scales are mentioned below, a 5-point Likert scale ranging from 1 = totally disagree to 5 =
totally agree was used. Children (age 6-18) have been found to prefer the Likert scale
over other scales (van Laerhoven, van der Zaag-Loonen, & Derkx, 2004) and to be unable
to grasp more than five response options (Chambers & Johnston, 2002; Streiner &
Norman, 2008).
Self-reported measures of behaviour were used because observing real-life
behaviour of large samples of people is prohibitively costly and would also be extremely
difficult in practice. Self-reporting is error-prone because respondents might be unable or
unwilling to accurately report their own behaviour (Thøgersen, 2008). To make it easier
to accurately report one’s behaviour, the individual behaviour items were made as
specific as possible. Respondents reported their behaviour on an 8-point scale from
“never” (coded as 1) to “seven times a week” (coded as 8) for each of eight items:
“Thinking of an ordinary week, how often do you eat fruit/vegetables as a part of your
breakfast/as part of your lunch/as part of your dinner/in-between meals”. Notice that we
asked for the frequency of consumption rather than specific portion sizes, which we
hoped decreased the risk of impression management (i.e., pushing respondents to
exaggerate how much they live up to the official guidelines). Further, we checked the
validity of the self-reported behaviour by also asking parents and children to report their
child’s/parents’ behaviour. The correlation between parents’ self-reported fruit and
vegetable intake and adolescents’ report of their behaviour is significant and positive (r =
148
.33) and the correlation between adolescents’ self-reported fruit and vegetable intake and
their parent’s report of their behaviour is even stronger (r = .46). It seems reasonable to
assume that the difference between these two correlations reflects the fact that parents are
able to report their children’s behaviour with greater certainty than vice versa. Be that as
it may, according to Cohen (1988), a correlation of .30/.50 is considered moderate/strong,
meaning that the obtained correlations support the validity of the self-reported behaviour
measures.
The adolescent’s perception of others’ behaviour (a descriptive norm) was
measured by three items: “My mum/dad/friends eat(s) a lot of fruit and vegetables”. The
adolescent’s perception of others’ attitudes (injunctive norm) was measured by the
following items: “My mum/dad/friends think(s), I should eat more fruit and vegetables”.
For these items, a “don’t know” option was added to the scale.
The items regarding parents’ behaviour and attitudes are assumed to capture
adolescents’ perception of the dominant family norms. It is likely that some adolescents
live in a context where their two parents differ in opinions and behaviour. However, a
Cronbach’s alpha of .85 suggests a rather coherent injunctive family norm (mum/dad
thinks…). Cronbach alpha for the descriptive family norm (mum/dad does…) is only .53
which suggests that this norm is not equally coherent, but it is still reasonable to assume
that it exists.
The respondent’s self-efficacy with regard to eating fruit and vegetables was
measured using nine items adapted from Perry et al.’s (2008) Physical Activity and
Healthy Food and Efficacy Scale for Children (PAHFE). The question was “How sure are
you that you can eat more fruit and vegetables….” with items focused on the situational
context such as: “…when watching TV or DVD” or “… when you are busy”. The
149
possible responses were “Not sure at all” (1), “Not too sure” (2), “Sure” (3), “Very sure”
(4), “Completely sure” (5). Cronbach’s alpha was .86 for children and .92 for parents.
The respondent’s outcome expectations (Bandura, 1986) were measured by
asking the respondent to evaluate seven possible outcomes of eating fruit and vegetables
such as: “I will like myself better if I eat more fruit and vegetables” or “I will lose weight
if I eat more fruit and vegetables”. Cronbach’s alpha was .81 for children and .82 for
parents.
Table 2
Overview over key variables
Study variables
Valid
Adolescents
Missing Mean Min.
Max.
Behaviour
“Thinking of an ordinary week, how often do you eat”:
…fruit as a part of
1300
188
2.64
1
8
your breakfast
…fruit as a part of
1293
195
4.11
1
8
your lunch
…fruit as a part of
1287
201
2.64
1
8
your dinner
…fruit in between
1299
189
5.03
1
8
meals
…vegetables as part
1302
186
1.40
1
8
of your breakfast
…vegetables as part
1302
186
3.55
1
8
of your lunch
…vegetables as part
1297
191
5.41
1
8
of your dinner
…vegetables in
1295
193
2.87
1
8
between meals
Self-efficacy
“How sure are you that you can eat more fruit and vegetables”:
…every day
1305
183
3.26
1
5
…when coming home 1296
192
3.17
1
5
from school/work
…when watching TV
1300
188
2.97
1
5
or DVD
…when sitting at the
1298
190
2.75
1
5
computer
…when your friends
1298
190
2.91
1
5
are around
…when you are bored 1295
193
3.22
1
5
…when you are in a
1297
191
2.52
1
5
bad mood
...when junk food is
1293
195
2.63
1
5
around
…when you are busy
1302
186
2.69
1
5
150
Valid
Parents
Missing Mean
Min.
Max.
779
709
3.25
1
8
774
714
3.97
1
8
756
732
2.19
1
8
785
703
5.77
1
8
773
715
1.32
1
8
775
713
4.95
1
8
779
709
6.46
1
8
776
712
3.30
1
8
777
780
711
708
3.48
3.54
1
1
5
5
778
710
3.22
1
5
777
711
2.95
1
5
778
710
3.04
1
5
776
776
712
712
3.04
2.74
1
1
5
5
778
710
2.87
1
5
778
710
2.86
1
5
Outcome expectations
“Please answer what will happen if you eat more fruit and vegetables”:
I will be in better
1289
199
3.73
1
5
775
shape, if I eat more
fruit and vegetables
I will like myself
1290
198
3.69
1
5
778
better, if I eat more
fruit and vegetables
I will get more
1288
200
4.09
1
5
774
energy, if I eat more
fruit and vegetables
I will lose weight, if I
1288
200
3.43
1
5
775
eat more fruit and
vegetables
I will be better
1287
201
3.43
1
5
776
looking, if I eat more
fruit and vegetables
If I eat more fruit and
1294
194
2.74
1
5
775
vegetables, so will the
rest of my family
My family will be
1295
193
342
1
5
775
pleased, if I eat more
fruit and vegetables
Descriptive norms
“Please think about how people you know eat”:
My mum eats a lot of
1182
306
3.94
1
5*
fruit and vegetables
My dad eats a lot of
1148
340
3.33
1
5*
fruit and vegetables
My friends eat a lot of
945
543
3.35
1
5*
fruit and vegetables
Injunctive norms
“Please answer what people you know think you should do”:
My friends think I
819
669
2.25
1
5*
should eat more fruit
and vegetables
My father thinks I
1035
453
2.87
1
5*
should eat more fruit
and vegetables
My mother thinks I
1114
374
3.23
1
5*
should eat more fruit
and vegetables
*A “Don’t know” option was added to the scale
713
3.56
1
5
710
3.78
1
5
714
4.05
1
5
713
3.75
1
5
712
3.68
1
5
713
3.47
1
5
713
2.43
1
5
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
6.3. Results
The analyses were conducted by means of structural equation modelling
(SEM) using AMOS 21 (Arbuckle, 2007). In SEM, the measurement model is a
confirmatory factor analysis (CFA) model and the theoretical constructs are latent factors
extracted from the manifest variables (Bagozzi, 1994). The key results from both the CFA
and the SEM are presented below in three steps.
151
6.3.1. Step 1
In Step 1 the effects of self-efficacy and outcome expectations on
adolescents’ behaviour were analysed while also including parents’ self-efficacy,
outcome expectations and behaviour in the analysis.
The bivariate correlations between the latent constructs based on CFA are
shown in Table 3. The fit indices show that the CFA model fits the data well. The
correlation matrix shows that all other latent variables correlate significantly with
adolescent behaviour, with adolescent self-efficacy, parent behaviour and adolescent
outcome expectations being the strongest predictors. The correlations reported in Table 3
are consistent with Hypothesis 1a, but some of them are inconsistent with Hypothesis 1b.
However, a multivariate analysis is necessary to properly test the hypotheses. This
analysis was conducted by means of SEM using all other variables included in the CFA
as predictors of adolescent behaviour, as is also shown in Table 3. The SEM identifies the
same variables as the strongest predictors of adolescents’ behaviour as the CFA. This
means that Hypothesis 1a is supported. However, since adolescent self-efficacy is a
stronger predictor of adolescents’ healthy eating than is parent behaviour, Hypothesis 1b
is not supported.
152
Table 3
The impact of adolescents’ and parents’ self-efficacy and outcome expectations and parents’ behaviour on
adolescent behaviour
Adolescent
behaviour
Adolescent
selfefficacy
Adolescent
outcome
expectations
Parent
behaviour
Parent
selfefficacy
Parent
outcome
expectations
R2
Correlations between latent variables
Adolescent
behaviour
Adolescent
self-efficacy
Adolescent
outcome
expectations
Parent
behaviour
Parent selfefficacy
Parent outcome
expectations
1.00
.57***
1.00
.37***
.28***
1.00
.41***
.18***
.09
1.00
.16***
.18***
.10*
.37***
1.00
.12*
.06
.22***
.20***
.14**
1.00
Structural modela
Adolescent
0.47***
0.23**
0.35***
-0.07
-0.01
0.45
behaviour
Parent
0.35*** 0.15**
0.16
behaviour
a
Standardized solution, only the structural model. Fit indices: Chi square = 342.117, 159 d.f., p < .001; CFI
= .96; RMSEA = .039 (90% confidence interval: .033-.045).
b
Based on CFA. The rest of the AMOS output from both analyses can be acquired from the first author. Fit
indices: Chi square = 335.637; 157 d.f., p < .001; CFI = .96; RMSEA = .039 (90% confidence interval:
.033-.045). * p < .05, ** p < .01, *** p < .001.
6.3.2. Step 2
In Step 2, the behavioural impact of dominant family norms as perceived by
the adolescent were investigated while also including the adolescent’s self-efficacy,
outcome expectations and parents’ (self-reported) behaviour. Since normative influence is
assumed to be mediated through the actor’s perceptions and interpretations (Grønhøj &
Thøgersen, 2012), adolescents’ perception of their parents’ behaviour was expected to be
more predictive of adolescent behaviour than was parents’ self-reported behaviour.
Table 4 shows the correlations between the latent constructs based on CFA.
Again, the fit indices show that the CFA model fits the data well. Adolescents’ behaviour
153
correlates significantly with their perception of parent behaviour, but contrary to our
expectations and Hypothesis 2, the correlation is weaker than with parent’s self-reported
behaviour.
Table 4 also presents the results of a SEM using all other variables included
in the CFA as predictors of adolescent behaviour. According to this analysis, adolescents’
perception of parent behaviour is not a significant predictor when these other variables
are included which means that Hypothesis 2 is not supported.
Table 4
The impact of adolescents’ perception of parents’ behaviour, parents’ self-reported behaviour and
adolescent’s self-efficacy and outcome expectations on adolescent behaviour
Adolescent
behaviour
Adolescent
selfefficacy
Adolescent
outcome
expectations
Adolescent
perception
of parent
behaviour
Parent
behaviour
R2
Correlations between latent variablesb
Adolescent behaviour
Adolescent selfefficacy
Adolescent outcome
expectations
Adolescent perception
of parent behaviour
Parent behaviour
1.00
.57***
1.00
.37***
.28***
1.00
.32***
.27***
.19**
1.00
.40***
.18***
.09
.33***
1.00
Structural modela
Adolescent behaviour
.46***
.22***
.09
.31***
0.43
Adolescent perception
.29**
0.09
of parent behaviour
a
Standardized solutions, only the structural model. Fit indices: Chi square = 307.058, 99 d.f., p < .001; CFI
= .93; RMSEA = .053 (90% confidence interval: .046-.059).
b
Based on CFA. The rest of the AMOS output from both analyses can be acquired from the first author. Fit
indices: Chi square = 260.776; 95 d.f., p < .001; CFI = .94; RMSEA = .048 (90% confidence interval: .041.055). * p < .05, ** p < .01, *** p < .001.
6.3.3. Step 3
In step 3, the impact of family and friends’ norms on adolescent behaviour
were compared while distinguishing between descriptive (perceived behaviour) and
154
injunctive (perceived attitude) norms. Since it was impossible to obtain measures of selfreported behaviour from friends, we only compared adolescents’ perceptions of parents’
and friends’ behaviour in Step 3, despite this measure being more weakly correlated with
adolescent behaviour than is parents’ self-reported behaviour in Step 2. Bivariate
correlations between the latent constructs based on CFA are shown in Table 5. The fit
indices show that the CFA model fits the data well. The correlation matrix shows that all
other latent variables correlate significantly with adolescents’ behaviour, with adolescent
self-efficacy, outcome expectations and the descriptive norms of parents as the strongest
predictors. As predicted, the descriptive norms of parents correlate more strongly with
adolescent behaviour than do the descriptive norms of friends. Surprisingly, injunctive
norms of both parents and friends correlate negatively and significantly with adolescent
behaviour. Still, the correlations are consistent with Hypotheses 3 and 4.
Table 5 also presents the results of the SEM using all other variables
included in the CFA as predictors of adolescent behaviour. When including the other
predictors, the descriptive norms of friends and the injunctive norms of both parents and
friends are not significant. With parents influencing adolescents more than friends,
Hypothesis 3 is supported. Hypothesis 4 is also supported since only the descriptive norm
(of parents) predicts adolescent behaviour.
155
Table 5
The impact of descriptive and injunctive norms regarding family and friends and adolescents’ own selfefficacy and outcome expectations on adolescent behaviour
Adolescent
behaviour
Adolescent
selfefficacy
Correlations between latent variablesb
Adolescent
1.00
behaviour
Adolescent self.56***
1.00
efficacy
Adolescent
outcome
.37***
.27***
expectations
Descriptive
.30***
.24***
norms, parents
Injunctive norms,
-.32***
-.31***
parents
Descriptive
.20***
.27***
norms, friends
Injunctive norms,
-.25***
-.14*
friends
Adolescent
outcome
expectations
Descriptive
norms,
parents
Injunctive
norms,
parents
Descriptive
norms,
friends
Injunctive
norms,
friends
R2
1.00
.21***
1.00
.09
.01
1.00
.21***
.12*
-.07
1.00
.09
-.13*
.66***
.13*
1.00
Structural modela
Adolescent
.39***
.26***
.14**
-.16
.02
-.10
0.44
behaviour
a
Standardized solutions, only the structural model. Fit indices: Chi square = 171.768, 60 d.f., p < .001; CFI
= .95; RMSEA = .050 (90% confidence interval: .041-.058).
b
Based on CFA. The rest of the AMOS output for both analyses can be acquired from the first author. Fit
indices: Chi square = 171.843, 61 d.f., p < .001; CFI = .95; RMSEA = .049 (90% confidence interval: .040.058). * p < .05, ** p < .01, *** p < .001.
The finding that adolescents are less influenced by friends than by parents
when it comes to healthy eating is bolstered by a simple analysis of adolescents’ ability to
answer questions about their mother/father/friends’ healthy eating behaviour and
expectations regarding the adolescent’s behaviour (descriptive and injunctive norms,
respectively). For these items, a “don’t know”3 response option was offered, and a
Due to the large proportion of “don’t know” responses, models with “don’t know” coded as
3
missing or as “either or” were compared, but without substantially different results. This indicates
that it is rather random whether the adolescent chooses to answer “either or” or “don’t know”.
Therefore the models with “don’t know” coded as “missing” were chosen.
156
substantially larger number of adolescents chose this option in response to questions
about their friends’ expectations and behaviour: 37.9% and 28.3% respectively, compared
to fathers (19% and 10%) and mothers (12.7% and 8.4%). This suggests that adolescents
are more uncertain about their friends’ than their parents’ expectations and behaviour,
probably because adolescents (irrespective of age group) are more exposed to parents’
healthy eating behaviour than to that of friends.
6.4. Discussion
The aim of this research was to compare the social influence of parents and
friends on adolescents’ healthy eating, focusing on fruit and vegetable intake, while
distinguishing between descriptive and injunctive norms and also including adolescent
self-efficacy and outcome expectations regarding fruit and vegetable intake. Previous
studies have not included all these factors in the same analysis. Consumer socialisation
theory suggests that the grip of parents gradually loosens in adolescence with friends
taking over (John, 1999). However, according to this study this is not the case for healthy
eating. The results presented here show that parents continue to be the main influence on
adolescents’ healthy eating behaviour whereas friends seem to have virtually no
influence, although this might partly depend on how we defined ‘healthy eating’ (as fruit
and vegetable intake rather than, e.g., cutting down on unhealthy food). Further, this
study shows that descriptive norms (what parents do) are more important than are
injunctive norms (what parents say) when it comes to healthy eating – the sort of eating
that parents would be expected to encourage. We found a surprising negative correlation
between injunctive norms and behaviour. We cannot say with certainty, but one might
157
speculate that this is due to others thinking that one should eat more fruit and vegetables
only in cases where one does not already eat a lot.4
Fitzgerald et al. (2013) found that peer support for healthy eating and selfefficacy were significantly stronger for boys than for girls. The fact that in every surveyed
age group of Danish adolescents, significantly more girls than boys eat fruit and
vegetables every day (Sundhedsstyrelsen, 2008), also suggests gender differences.
However, the present study did not find significant differences regarding self-efficacy,
outcome expectations, social norms or behaviour between genders, nor between age
groups.5
As with all other studies, this one has its limitations. First of all, this is a
cross-sectional study conducted in one country at a single point in time. Both the cultural
context and changes over time are important aspects of socialisation processes, so future
research should include both cross-cultural and longitudinal studies. As has been pointed
out by de Castro (1999), genetics and heredity are further factors that might account for
parent-child similarities in food intake, in addition to socialisation and social influence.
However, the present study has no basis for saying anything about the importance of
these factors. Another limitation is the measurement of behaviour (which could in
principle be observed) by self-report, which may be inaccurate. As was mentioned in the
method section, respondents may be unable or unwilling to accurately report their own
behaviour (Thøgersen, 2008). In the method section we also reported a number of
measures that we took to reduce the errors in self-reports; significant, positive and quite
4
We are grateful to an anonymous reviewer for suggesting this interpretation.
5
The analyses may be acquired from the first author.
158
strong correlations between self-reported behaviour and reports by the respondent’s
parent or child support the validity of the self-report measures.
On the basis of SCT and previous findings (e.g. Baker et al., 2000), it was
expected that adolescent behaviour would be more strongly related to their perception of
parents’ behaviour than to parents’ self-reported behaviour. There could be several
reasons why this was not what we found. One of these is the fact that the construct
reliability of adolescents’ perception of parents’ behaviour (also referred to as descriptive
family norms) is weaker than that of all other predictor variables. The weak reliability
might have led to its relationship with other variables being attenuated. Another reason
might be that some unaccounted for “third variable” increases the correlation between
parent and child self-reported behaviour, over and above the effects of social influence.
For example, it is possible that the strong correlation between children’s and parents’
eating behaviour is partly due to genetic factors (de Castro, 1999). Perhaps even more
obviously, the correlation between the two self-reported behaviours might partly be due
to a shared family context containing shared socio-structural factors that either facilitate
or impede fruit and vegetable intake for all family members. For instance, previous
studies have found that availability at home is an important predictor of fruit and
vegetable intake (Kristjánsdóttir et al., 2009; Neumark-Sztainer, Wall, Perry, & Story,
2003; Young et al., 2004). Hence, future research on this topic should both aim for a more
reliable measure of descriptive family norms, as perceived by the children, and develop
ways to measure how facilitating or impeding the family context is for the analysed
behaviour. And until this unexpected finding has been replicated in other studies, it would
be prudent not to draw any practical implications from it.
The latter reservation is relevant for other results of this study that await
support from other studies as well. However, many results of this study are in line with
159
much contemporary research in related areas, which increases their face validity. For
example, the observed impact of self-efficacy and outcome expectations on adolescents
supports previous research recommending that healthy eating interventions targeted at
adolescents should aim at strengthening their self-efficacy and clearly emphasise positive
outcomes. Also, the observed strength of parental influence compared to that of friends
suggests that the family context should be included when conducting healthy eating
interventions and that parents – as gatekeepers – should be made aware of their role
model influence with regard to healthy eating. It is important to stress, however, that we
are not implying, for example, that school interventions should be de-emphasised. The
present study is mute about the effectiveness of such interventions. The stronger effect of
descriptive than injunctive norms shows that it is not enough for parents to preach healthy
eating; they have to demonstrate it with their own good example. In sum, active parent
involvement is recommended when conducting healthy eating interventions among
adolescents (see also Pedersen, Grønhøj, & Bech-Larsen, 2012).
According to this study, friends’ attitudes and behaviour do not influence
adolescents’ fruit and vegetable intake which conflicts with previous research findings,
for example, that adolescents conform to a common eating norm thereby creating a group
feeling (Stead et al., 2011). It is a limitation of this study that the adolescents were asked
questions about their “friends” in general which demands that they aggregate over several
persons with perhaps very different attitudes and behaviour. This might explain why the
adolescents in this study found it difficult to answer questions about their friends’
attitudes and behaviour. However, this difficulty might also be due to eating fruits and
vegetables not being an identity-defining priority among adolescents. If this finding can
be replicated in other studies, it suggests that healthy eating interventions targeting groups
160
of friends should not necessarily rely on friends influencing each other, but perhaps
instead aim to create a group feeling about healthy eating.
In conclusion, this study has produced new knowledge about the role of
important predictors on healthy eating among adolescents: self-efficacy, outcome
expectations and not least descriptive and injunctive norms of parents and friends. Future
research should build on these findings by exploring how interventions can build up selfefficacy and favourable outcome expectations while also including parents as positive
models as a means to increase healthy eating among adolescents.
Acknowledgements
This study is part of the research project “Step by step changes of children’s
preferences towards healthier food’’ which was funded by the Danish Ministry of
Science, Technology and Innovation, grant no. 09/061357. The authors would like to
thank student assistants Jacob Heiss Rosendahl, Astrid Refsgaard, Ken Jørgensen and
Christina Bæklund for their work in relation to the data collection and Birgitte Steffensen
for proofreading the manuscript. We are also grateful for constructive comments on a
previous version of the manuscript from C. Peter Herman and two anonymous reviewers.
161
References
Arbuckle, J. L. (2007). AMOS 18 User's guide. Chicago: SPSS Inc.
Bagozzi, R. P. (1994). Structural equation models in marketing research: Basic principles.
In R. P. Bagozzi (Ed.), Principles of marketing research (pp. 317-385).
Cambridge: Blackwell Publishers.
Baker, C. W., Whisman, M. A., & Brownell, K. D. . (2000). Studying intergenerational
transmission of eating attitudes and behaviors: Methodological and conceptual
questions. Health Psychology, 19(4), 376-381. doi: 10.1037/0278-6133.22.2.189
Ball, K., Jeffery, R., Abbott, G., McNaughton, S., & Crawford, D. . (2010). Is healthy
behavior contagious: Associations of social norms with physical activity and
healthy eating. International Journal of Behavioral Nutrition and Physical
Activity, 7(1). doi: 10.1186/1479-5868-7-86
Bandura, Albert. (1977). Social learning theory (1st ed.). Englewood Cliffs, NJ: PrenticeHall.
Bandura, Albert. (1986). Social foundations of thought and action: A social cognitive
perspective. Englewood Cliffs, NJ: Prentice-Hall.
Bandura, Albert. (2004). Health promotion by social cognitive means. Health Education
& Behavior 31(2), 143-164.
Bauer, K W, Laska, Melissa N, Fulkerson, Jayne A, & Neumark-Sztainer, D. (2011).
Longitudinal and secular trends in parental encouragement for healthy eating,
physical activity and dieting throughout the adolescent years. Journal of
Adolescent Health, 49(3), 306. doi: 10.1016/j.jadohealth.2010.12.023
Birch, L., & Fisher, J. (1998). Development of eating behaviors among children and
adolescents. Pediatrics, 101(3), 539-549.
Brislin, Richard W. (1970). Back-translation for cross-cultural research. Journal of CrossCultural Psychology, 1(3), 185-216. doi: 10.1177/135910457000100301
Bruening, M. , Eisenberg, M. , MacLehose, R. , Nanney, M. , Story, M. , & NeumarkSztainer, D. (2012). Relationship between adolescents’ and their friends’ eating
behaviors: Breakfast, fruit, vegetable, whole-grain and dairy intake. Journal of
Academy of Nutrition and Dietetics, 112, 1608-1613. doi:
10.1016/j.jand.2012.07.008
Chambers, C. T., & Johnston, C. (2002). Developmental differences in children's use of
rating scales. Journal of Pediatric Psychology, 27(1), 27-36.
162
Chan, Kara, Prendergast, Gerard, Grønhøj, Alice, & Bech-Larsen, Tino. (2010). The role
of socializing agents in communicating healthy eating to adolescents: A cross
cultural study. Journal of International Consumer Marketing, 23(1), 59–74. doi:
10.1080/08961530.2011.524578
Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of normative
conduct: A theoretical refinement and re-evaluation of the role of norms in human
behavior. Advances in Experimental Social Psychology, 24, 201-234. doi:
10.1016/S0065-2601(08)60330-5
Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative
conduct: Recycling the concept of norms to reduce littering in public places.
Journal of Personality and Social Psychology, 58, 1015-1026. doi:
10.1037//0022-3514.58.6.1015
Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity and
compliance. In D. T. Gilbert, S. T. Fiske & G. Lindzey (Eds.), The handbook of
social psychology (4 ed., Vol. 2, pp. 151–192 ). New York: McGraw-Hill.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.).
Hillsdale NJ: Erlbaum.
Croll, J. K. , Neumark-Sztainer, D., & Story, M. (2001). Healthy eating. What does it
mean to adolescents? Journal of Nutrition Education (33), 193-198.
de Castro, J. M. (1999). Behavioral genetics of food intake regulation in free-living
humans. Nutrition, 15(7-8), 550-554. doi: 10.1016/S0899-9007(99)00114-8
Fitzgerald, A., Heary, C., Kelly, C., Nixon, E., & Shevlin, M. (2013). Self-efficacy for
healthy eating and peer support for unhealthy eating are associated with
adolescents’ food intake patterns. Appetite, 63, 48-58. doi:
10.1016/j.appet.2012.12.011
Geller, K., & Dzewaltowski, D. (2010). Examining elementary school-aged children's
self-efficacy and proxy efficacy for fruit and vegetable consumption. Health
Education & Behavior, 37(4), 465-478. doi: 10.1177/1090198109347067
Gitelson, I. , & McDermott, D. (2006). Parents and their young adult children: Transitions
to adulthood. Child Welfare, 86(5), 853-866.
Grønhøj, Alice, & Thøgersen, John. (2012). Action speaks louder than words: The effect
of personal attitudes and family norms on adolescents' pro-environmental
behaviour. Journal of Economic Psychology, 33(1), 292-302. doi:
10.1016/j.joep.2011.10.001
163
Herman, C. Peter, Roth, Deborah A, & Polivy, Janet. (2003). Effects of the presence of
others on food intake: A normative interpretation. Psychological bulletin, 129(6),
873-886. doi: 10.1037/0033-2909.129.6.873
Howland, M. , Hunger, J. , & Mann, T. (2012). Friends don’t let friends eat cookies:
Effects of restrictive eating norms on consumption among friends. Appetite, 59,
505-509. doi: 10.1016/j.appet.2012.06.020
John, Deborah Roedder. (1999). Consumer socialization of children: A retrospective look
at twenty-five years of research. Journal of Consumer Research, 26(3), 183-213.
Kristjánsdóttir, A. G., De Bourdeaudhuij, I., Klepp, K.-I., & Thorsdóttir, I. (2009).
Children’s and parents’ perceptions of the determinants of children’s fruit and
vegetable intake in a low-intake population. Public Health Nutrition, 12(8), 12241233. doi: 10.1017/S1368980008004254
Lake, A. A., Rugg-Gunn, Andrew J., Hyland, Rob M., Wood, Charlotte E., Mathers, John
C., & Adamson, Ashley J. (2004). Longitudinal dietary change from adolescence
to adulthood: Perceptions, attributions and evidence. Appetite, 42(3), 255-263.
doi: 10.1016/j.appet.2003.11.008
Lally, P., Bartle, N., & Wardle, J. (2011). Social norms and diet in adolescents. Appetite,
57, 623–627. doi: 10.1016/j.appet.2011.07.015
Lau, Richard R., Quadrel, Marilyn Jacobs, & Hartman, Karen A. (1990). Development
and change of young adults' preventive health beliefs and behavior: Influence
from parents and peers. Journal of Health and Social Behavior, 31(3), 240-259.
McClain, A. D., Chappuis, C., Nguyen-Rodriguez, S. T., Yaroch, A. L., & Spruijt-Metz,
D. (2009). Psychosocial correlates of eating behavior in children and adolescents:
A review. International Journal of Behavioral Nutrition and Physical Activity,
6(54). doi: 10.1186/1479-5868-6-54
Neumark-Sztainer, D., Wall, M. , Perry, C., & Story, M. (2003). Correlates of fruit and
vegetable intake among adolescents: Findings from Project EAT. Preventive
Medicine, 37(3), 198-208. doi: 10.1016/S0091-7435(03)00114-2
Pedersen, Susanne, Grønhøj, Alice, & Bech-Larsen, Tino. (2012). Family members' roles
in healthy-eating socialization based on a healthy-eating intervention. Young
Consumers: Insight and Ideas for Responsible Marketers, 13(3), 208-223. doi:
10.1108/17473611211261610
Perry, Christina M. , De Ayala, R. J., Lebow, Ryan, & Hayden, Emily. (2008). A
validation and reliability study of the physical activity and healthy food efficacy
scale for children. Health Education & Behavior, 35.
164
Rasmussen, Mette, Krolner, Rikke, Klepp, Knut-Inge, Lytle, L, Brug, Johannes, Bere, E,
& Due, Pernille. (2006). Determinants of fruit and vegetable consumption among
children and adolescents: A review of the literature. Part I: quantitative studies. Int
J Behav Nutr Phys Act, 3, 22.
Resnicow, K., Davis-Hearn, M., Smith, M., Baranowski, T., Lin, L. S., Baranowski, J., . .
. Wang, D. T. (1997). Social-cognitive predictors of fruit and vegetable intake in
children. Health Psychology, 16(3), 272-276.
Salvy, Sarah-Jeanne, de la Haye, K. , Bowker, J. C. , & Hermans, R. C. J. (2012).
Influence of peers and friends on children's and adolescents' eating and activity
behaviors. Physiology & Behavior, 106, 369-378. doi:
10.1016/j.physbeh.2012.03.022
Stead, M., McDermott, L., MacKintosh, A. M., & Adamson, A. . (2011). Why healthy
eating is bad for young people’s health: Identity, belonging and food. Social
Science & Medicine, 72, 1131-1139. doi: 10.1016/j.socscimed.2010.12.029
Streiner, D. L., & Norman, G. R. (2008). Health measurement scales: A practical guide
to their development and use (4th ed.). New York: Oxford University Press.
Sundhedsstyrelsen. (2008). Undersøgelse af 11-15 åriges livsstil og sundhedsvaner 19972006 [A study of 11-15 year-olds’ life styles and health habits 1997-2006]
Copenhagen Sundhedsstyrelsen [Danish Health and Medicines Authority].
Thøgersen, John. (2008). Social norms and cooperation in real-life social dilemmas.
Journal of Economic Psychology, 29(4), 458-472. doi: 10.1016/j.joep.2007.12.004
van Laerhoven, H., van der Zaag-Loonen, H. J., & Derkx, B. H. F. (2004). A comparison
of Likert scale and visual analogue scales as response options in children’s
questionnaires. Acta Paediatrica, 93(6), 830-835. doi:
10.1080/08035250410026572
WHO. (2000). Obesity: Preventing and managing the global epidemic. Report of a WHO
Consultation. WHO Technical Report Series 894. Geneva: WHO.
WHO. (2003). Diet, nutrition and the prevention of chronic diseases. Geneva: WHO
WHO. (2004). Young people's health in context. Health behaviour in school-aged
children (HBSC): International report from the 2001/2002 survey. In C. R. C.
Currie, A. Morgan, R. Smith, W. Settertobulte, O. Samdal, V. B. Rasmussen
(Ed.), Health Policy Series: Health for Children and Adolescents. Copenhagen:
WHO Regional Office for Europe.
165
Young, E., Fors, S. W., Fasha, E., & Hayes, D. (2004). Associations between perceived
parent behaviors and middle school student fruit and vegetable consumption.
Journal of Nutrition Education and Behavior (36), 2-12.
166
7. TEXTING YOUR WAY TO HEALTHIER EATING? EFFECTS OF
PARTICIPATING IN A FEEDBACK INTERVENTION USING TEXT
MESSAGING ON ADOLESCENTS’ FRUIT AND VEGETABLE
INTAKE1
Abstract
This study investigates the effects of a feedback intervention employing text
messaging during 11 weeks on adolescents’ behavior, self-efficacy and outcome
expectations regarding fruit and vegetable intake. A pre- and post-survey was completed
by 1488 adolescents randomly allocated to a control group and two experimental groups.
One experimental group set weekly goals on F/V intake, reported their consumption daily
and subsequently received feedback on their performance via mobile text messaging
(SMS). The second experimental group received the same treatment and, in addition,
received a 45-minute nutrition education session from a dietician during school. The
direct effects of the interventions were not significant. However, for adolescents
participating in the SMS routines, there were significant effects of the level of activity in
the intervention, reflected in the number of sent text messages. Participants sending more
than half of the possible text messages significantly increased their fruit and vegetable
intake. Participants sending between 10 and 50% of the possible text messages
experienced a significant drop in self-efficacy and those sending less than 10%
experienced a significant drop in outcome expectations. The findings suggest that
1
This chapter is submitted for review as: Pedersen, S., Grønhøj, A. & Thøgersen, J. Texting your
way to healthier eating? Effects of participating in a feedback intervention using text messaging
on adolescents’ fruit and vegetable intake.
167
participants’ active engagement in an intervention is crucial to its success. Implications
for health-promoting interventions are discussed.
7.1 Introduction
Most adolescents do not eat the recommended daily amount of fruit and
vegetables (Lynch et al., 2014; Rasmussen et al., 2006) which can have negative health
consequences, including chronic diseases and obesity (Bazzano, 2006; WHO, 2003).
However, changing adolescents’ eating behavior to become healthier is – as most parents
can confirm – a challenging task.
The end goal of health interventions is usually behavioral change, but
positive changes in important antecedents of behavior, such as self-efficacy or behavioral
intentions, are important steps on the way (Baranowski, Lin, Wetter, Resnicow, & Hearn,
1997). Hence, it is important to identify key mediators of behavioral impacts (Cerin,
Barnett, & Baranowski, 2009).
Text messaging via mobile phones’ Short Message Service (SMS) has been
suggested as an effective approach to promote health behavior among adolescents (Bauer,
de Niet, Timman, & Kordy, 2010; Bech-Larsen & Grønhøj, 2013; Shapiro et al., 2008;
Woolford, Clark, Strecher, & Resnicow, 2010). Text messaging is a two-way tool and can
hence be used by participants to report on a specific behavior and receive response
(feedback) that is tailored to their reports. This interactivity opens opportunities for more
active involvement by the participants than the common one-way information
interventions, including much nutrition education in a school setting (Birnbaum, Lytle,
Story, Perry, & Murray, 2002; Story et al., 2000). However, there has been little attention
to how successful such interventions are with regard to creating active participation and
168
what that means for the impact of the intervention, both in general and specifically with
regard to interventions targeting adolescents. On this background, this study’s
contribution is twofold: to provide knowledge on (1) the direct and indirect effects on
adolescents’ fruit and vegetable intake through a feedback intervention delivered by text
messaging and (2) the importance of participants’ level of activity in the intervention for
the intervention’s success.
Text messaging is one of (pre)adolescents’ preferred communication tools
(Davie, Panting, & Charlton, 2004; Faulkner & Culwin, 2005; Phau & Teah, 2009), with
girls using text messaging more than boys, according to some studies (Watten, Kleiven,
Fostervold, Fauske, & Volden, 2008). When considering, in addition, that it is easy to
use, accessible, affordable and available to all social groups (Fjeldsoe, Marshall, &
Miller, 2009) and makes it possible to receive and send brief, private, and personalized
messages or reminders (Preston, Walhart, & O'Sullivan, 2011), text messaging seems an
obvious choice for an interactive feedback intervention targeting (pre)adolescents. Also,
Shapiro et al. (2008) found that children (age 5–13) who engaged in text messaging,
compared to paper diaries, had lower attrition and significantly greater adherence to selfmonitoring. However, despite its many advantages, only few health intervention studies
targeting adolescents have used text messaging.
Several studies have documented the importance of self-efficacy for dietary
behavioral change (Cerin et al., 2009), including increasing adolescents’ fruit and
vegetable intake (e.g. Bere & Klepp, 2004; Fitzgerald, Heary, Kelly, Nixon, & Shevlin,
2013; Geller & Dzewaltowski, 2010; Young, Fors, Fasha, & Hayes, 2004). Self-efficacy
is the belief “that one has the power to produce desired changes by one’s actions”
(Bandura, 2004, p. 144). Personal and timely feedback on behavior via information or
169
exercises has been found to be an effective way to make participants reflect on their own
behavior (Schultz, 1999) and enhance their self-efficacy (Kluger & DeNisi, 1996;
Luszczynska, Tryburcy, & Schwarzer, 2007).
Outcome expectations – another key construct in Bandura’s (1986) Social
Cognitive Theory (SCT) – have also consistently been associated with dietary behavioral
change (Cerin et al., 2009), including increasing fruit and vegetable intake (Cerin et al.,
2009; Resnicow et al., 1997). Outcome expectations can be physical, social and/or selfevaluative expectations regarding the outcome of an anticipated behavior (Bandura,
1977). Feedback about the outcomes of one’s actions is an important input for forming
expectations about results of working towards a goal. Feedback is also a way to reduce
the effects of the delay between action and result (Kluger & DeNisi, 1996).
A more common approach than feedback intervention using text messaging
to promote healthy behavior among children and adolescents is health education that
focuses on nutrition information (e.g. Contento, Randell, & Basch, 2002). In practice, it
seems relevant to combine the two approaches. Indeed, it is often recommended to use
multi-component interventions for health behavior change interventions among
adolescents (Birnbaum et al., 2002). Hence, this study also tests whether the effects of a
feedback intervention can be enhanced by adding traditional health education.
Girls have been found to use their mobile phones more for text messaging
than boys (Watten et al., 2008). This, combined with the fact that in every surveyed age
group of Danish adolescents, significantly more girls than boys eat fruit and vegetables
every day (Sundhedsstyrelsen, 2008), suggest that there might be gender differences in
the effects of the tested interventions.
170
In sum, the aim of the present study was to investigate the impact of an interactive
text messaging-based feedback intervention on adolescents’ fruit and vegetable intake,
possibly mediated through increased self-efficacy and more favorable outcome
expectations, and to which extent these desirable impacts are contingent on additional
health education and participants displaying a certain level of activity in the intervention.
7.2. Hypotheses
Hypothesis 1a: Receiving and sending text messages about their fruit and
vegetable intake increases adolescents’ fruit and vegetable consumption and their related
self-efficacy and outcome expectations.
Hypothesis 1b: Receiving nutrition education from a dietician, in addition
to receiving and sending text messages about their fruit and vegetable intake, further
improves adolescents’ fruit and vegetable intake and their related self-efficacy and
outcome expectations, compared to only receiving and sending text messages.
Hypothesis 2: The more active participants are in the intervention, as
reflected in the proportion of possible text messages sent, the more behavior, self-efficacy
and outcome expectations will change in the intended direction.
Hypothesis 3: Girls will be more active in an intervention promoting fruit
and vegetable intake and will improve their behavior, self-efficacy and outcome
expectations significantly more than boys.
171
7.3. Method
7.3.1. Participants
Pupils from 5th and 9th grade were recruited class-wise in the Central
Denmark Region in September 2010. A sample of schools that had not participated in
health projects the last three years was drawn from the region’s school register so as to
give a broad representation of the region’s full curriculum of public schools (i.e.,
excluding private schools and schools that do not offer the oldest classes) in terms of
school size and geographical distribution, pupils’ socio-demographic background,
availability of school cafeteria and permission for pupils to leave the school area during
school time. Permissions to conduct the study were obtained from school principals (17
schools) and from class teachers (70 classes). Also, pupils and their parents received
detailed information about the project and it was made clear that participation was
voluntary and that participants were free to stop at any time. It was also emphasized that
the registration of consumption and goal data would be completely automatic and
anonymous and that the costs of the text messages (approximately 10 DKK or 2 US$)
would not be refunded. The pupils got further, detailed information from the research
assistants about the research project’s elements and duration and the opportunity to ask
further questions in class. Hence, they were assumed to be capable of giving informed
consent (France, 2004).2
2
There is no requirement for IRB approval in Danish universities, but the researchers nevertheless
made sure that the project complied with all existing ethical guidelines for this kind of social
science research.
172
7.3.2. Interventions
In order to avoid the possibility that pupils receiving different treatment
would influence one another and because the session with a dietician was in the class,
whole schools were randomly assigned to one of the three different treatments: (1) text
messaging, (2) text messaging + nutritional education and (3) a minimal intervention
control group. The intervention lasted for 11 weeks, including five weeks of text
messaging. This is a relatively brief intervention; however, in order to avoid excessive
attrition, brief interventions are usually recommended for interventions targeting children
(Stice, Shaw, & Marti, 2006). As a means to further reduce attrition, a prize (a trip to a
local theme park worth approx. 600 US$) was offered to the class sending the most text
messages.
Everyone answered a survey before and after the intervention. In addition,
the SMS and SMS/education groups were requested to report their daily fruit and
vegetable intake via text messaging every other week (five weeks in total) and received
automatic feedback on their intake after each report. Finally, the SMS/education group
also received a 45-minute education session from a dietician.
Classes in the SMS/education group were visited once by a dietician from
Centre for Public Health, Central Denmark Region who gave them a 45-minute session
about the importance of eating fruit and vegetables, following the official health
recommendations (6 portions of fruit and vegetables per day). The purpose was to expose
participants in this condition to a common type of school education on nutrition. Included
in this session was also a practical exercise where the pupils, in groups, were asked to
choose what they believed to be 600 grams from five kilos of mixed fruits and vegetables.
The chosen amounts were weighed and it was calculated how much they differed from
600 grams, thereby visualising both their own estimates and the recommended 600 grams
173
of fruit and vegetables for them. The content of the education sessions was coordinated
by the first author to ensure that all classes received identical information. This was
confirmed by audio recordings of the sessions.
Participants in the intervention groups received instruction from research
assistants about setting goals and reporting their intake. As recommended by Kluger and
Denisi (1996) and Locke et al. (1981), participants in the intervention groups were asked
to set ambitious, yet realistic goals for their weekly fruit and vegetable intake. In order to
maximize adherence and performance, these goals could be adjusted every other week.
Research assistants visited the classes twice: at the beginning and at the end
of the intervention, and a third time with the dietician in the SMS/education group.
Research assistants were instructed thoroughly before visits to the classes in order to
assure that the same information was given to the participants, and each visit was
evaluated in terms of observation notes and digital sound recordings.
Fruit and vegetable intake was – after a pre-test with five adolescents age 11
to 15 – considered self-measureable for adolescents. The counting of units equivalent to
100 grams of fruit and vegetables was facilitated by a counting table (a laminated A4
page with photos and verbal depictions of unit equivalents of different types and servings
of fruit and vegetables; see appendix A).
A computer program was specifically developed for this study for managing
the two-way text messaging which consisted of reporting by the participants and
automatic feedback following each report. Every night at 8 o’clock during the reporting
weeks, participants were automatically requested to send a report covering that day’s
units (for instance 2 f, 3 v) by a text message from the computer program. Immediately
after sending their report, participants received feedback from the computer program
174
comparing their daily report to the self-set weekly goal. The feedback was factual and
referred to the calculated differences between goal and intake (e.g., “You have now eaten
6 f and 8 v. In order to reach your weekly goal, you still need 14 f and 15 v before Sunday
evening”). Using concrete terms was assumed to facilitate adherence and performance
(Bech-Larsen & Grønhøj, 2013; Kluger & DeNisi, 1996).
A leaflet with questions and answers in case of technical problems was
handed out to participants and a hotline was available every day until midnight. A mobile
phone was made available to those who did not own one (only six pupils).3
7.3.3. The survey
The purpose of the survey was to measure the effects of the intervention.
The pre and post surveys were therefore identical except that the post-intervention survey
did not include socio-demographics. Because children (6–18) have been found to prefer
the Likert scale over other scales (van Laerhoven, van der Zaag-Loonen, & Derkx, 2004)
and to often be unable to grasp more than five response options (Chambers & Johnston,
2002; Streiner & Norman, 2008), a 5-point Likert scale ranging from 1 = totally disagree
to 5 = totally agree was used as much as possible.
For pre-testing the survey instrument, a school gave access to four pupils
(age 11) who completed the survey and afterwards discussed wording and scales with the
first author. Following adjustments, four new pupils from the same school went through
the same pre-test. A third pre-test was conducted with 30 adolescents (age 10–16), based
3
The dispersion of mobile phones in Denmark is close to 100% in the younger age groups, so this
(low) number was to be expected (Statistik, 2011).
175
upon which frequencies and scale reliabilities were checked. The self-efficacy scale was
translated from English to Danish and then back-translated (Brislin, 1970). Multi-item
scales were created by averaging the scores on individual items.
The participants’ behavior was measured with eight intake frequency items
(response options from 0 = “never” to 7 = “seven times a week”), four for fruit and four
for vegetables, referring to how often fruit and vegetables are consumed at each meal or
snack in an average week (i.e., breakfast, lunch, dinner, and between meals, see Table 1).
The scale had acceptable construct reliability (Cronbach’s alpha .75 [pre] and .74 [post]).
The respondents’ self-efficacy with regard to eating fruit and vegetables was
measured with nine items adapted from Perry et al.’s (2008) Physical Activity and
Healthy Food and Efficacy Scale for Children (PAHFE). The possible responses were
“Not sure at all” (1), “Not too sure” (2), “Sure” (3), “Very sure” (4), “Completely sure”
(5). The scale had excellent construct reliability (Cronbach’s alpha .87 [pre] and .89
[post]).
The respondents’ outcome expectations (Bandura, 2001, 2004) were
measured by an instrument adapted from Thøgersen and Grønhøj (Thøgersen & Grønhøj,
2010), asking participants to evaluate seven desirable outcomes of eating (more) fruit and
vegetables. The scale had excellent construct reliability (Cronbach’s alpha .82 [pre] and
.85 [post]).
7.3.4. Statistical analyses
Multiple regression analysis using SPSS 21 is used to test the hypotheses
about changes in behavior, self-efficacy and outcome expectations in relation to fruit and
176
vegetable intake from before to after the intervention. In all analyses, the baseline value
of the dependent variable is included among the predictors in order to control for the
generally observed regression towards the mean (Cohen, Cohen, West, & Aiken, 2003).
Hypotheses about interactions between gender and activity level were tested by means of
hierarchical regression analysis.
In order to determine the effects of variations in text messaging activity
level on changes in behavior, self-efficacy and outcome expectations related to fruit and
vegetable intake, participants in the two intervention groups were divided into three levels
of activity. Two cut-off points for the proportion of possible text messages (10% and
50%) sent during the five weeks were chosen based on an inspection of the distribution in
text messaging activity. Based on these cut-off points, three levels of activity were
defined: Low (below 10%; i.e., 0–3 text messages), medium (between 10 and 50%; i.e.,
4–17 text messages), and high (above 50%; i.e., 18–35 text messages).
7.4. Results
A total of 1488 pupils from 5th and 9th grade (mean age 10.9 and 14.8 years,
respectively) were allocated to a minimal intervention control group (n = 502), an SMSonly group (n = 489) or an SMS and nutrition education group (n=497). The three
experimental groups did not differ significantly on any of the included background
variables, reported in Table 1, at the Bonferroni corrected level of .01.
177
Table 1
Background information about participants based on socio-demographic data from the pre-intervention
survey
Total
Text messaging
Text messaging
+ nutritional
education
1488
489
497
502
50/50
50/50
48/52
53/47
50/50
50/50
51/49
49/51
boys/girls in 5th grade (%)
49/51
49/51
46/54
53/47
boys/girls in 9th grade (%)
52/49
12.9 (2.031)
52/48
12.8 (2.041)
49/51
12.9 (2.015)
53/47
12.9 (2.041)
10.9 (0.480)
10.9 (0.480)
10.9 (0.472)
10.9 (0.487)
9 graders
14.8 (0.482)
14.8 (0.460)
14.9 (0.466)
14.8 (0.513)
Parents (n)
781
291
244
246
15/85
17/83
16/84
12/88
7th –10th grade
8.9
5.6
10.2
11.4
High school or similar
7.4
9.8
7.1
4.9
Vocational education
Short further education (up to two
years)
College degree completed (2–4 years)
27.7
10.9
26.5
11.2
29.1
11.2
27.6
10.3
35.2
34.4
34.2
37.3
Graduate school (Masters, Doctorate
or equivalent)
Monthly household income before tax
9.9
12.6
8.2
8.6
0.7
0.7
0.4
0.9
8,000-1,4999 DKK /$1441–$2702
15,000-29,999 DKK/$2,702–$5,404
3.4
13.7
3.7
13.6
3.4
10.6
3.1
17.2
30,000-49,999 DKK/$5,404–$9,007
29.1
24.5
32.2
31.3
50,000-69,999 DKK/$9,007–$12,609
35.5
34.4
36.4
35.7
70,000 DKK or more/$12609 or more
17.7
23.1
16.9
11.9
Adolescents (n)
Gender
male/female (%)
Class level
5th/9th graders (%)
Mean age (st. dev.)
5th graders
th
Control group
Gender
male/female (%)
Parents’ educational level (%)
Less than 8,000 DKK/$1,441
Everyone was asked to complete a survey before and after the intervention
at school under the supervision of a research assistant. The response rate in the first
survey was 88% (range 64–97% for individual schools), and in the second survey also
88% (range 75–96%). Table 2 provides an overview of the behavioral and psychological
measures and means for both the pre- and post-intervention surveys in the three groups.
178
Table 2
Key variables: Mean item and scale scores for behavior, self-efficacy and outcome expectations in the three
experimental groups and at the three activity levels in the intervention groups
Control
group
Mean
Presurvey
Postsurvey
Text
messaging
Presurvey
Postsurvey
Text
messaging +
nutrition
education
Presurvey
< 10%
10 - 50%
> 50%
Postsurvey
PreSurve
y
Postsurvey
Presurvey
Postsurvey
Presurvey
Postsurvey
Behavior. “Thinking of an ordinary week, how often do you eat”: (1–8)
…fruit as a part of your
breakfast
…fruit as a part of your
lunch
…fruit as a part of your
dinner
…fruit in between meals
…vegetables as part of
your breakfast
…vegetables as part of
your lunch
…vegetables as part of
your dinner
…vegetables between
meals
Behavior scale average
2.74
2.66
2.68
2.75
2.48
2.61
2.47
2.46
2.67
2.59
2.58
2.84
4.06
4.27
4.27
4.13
4.00
4.11
4.07
3.95
3.95
3.80
4.26
4.38
2.75
2.46
2.65
2.53
2.49
2.46
2.45
2.52
2.66
2.51
2.58
2.47
5.15
5.03
4.98
5.05
4.95
4.79
4.68
4.61
4.84
4.75
5.17
5.18
1.46
1.41
1.37
1.37
1.38
1.31
1.47
1.27
1.39
1.43
1.33
1.34
3.47
3.36
3.65
3.58
3.54
3.35
2.99
2.97
3.60
3.34
3.86
3.79
5.38
5.17
5.43
5.40
5.42
5.22
4.98
4.82
5.59
5.28
5.54
5.58
2.89
2.71
2.91
2.75
2.82
2.67
2.61
2.50
2.96
2.58
2.92
2.89
3.50
3.39
3.50
3.45
3.39
3.32
3.23
3.11
3.45
3.28
3.53
3.57
Self-efficacy. “How sure are you that you can eat more fruit and vegetables”: (1–5)
…every day
…when coming home
from school/work
…when watching TV or
DVD
…when sitting at the
computer
…when your friends are
around
…when you are bored
…when you are in a bad
mood
...when junk food is
around
…when you are busy
Self-efficacy scale
average
3.23
3.19
3.30
3.20
3.27
3.14
3.07
2.91
3.41
3.09
3.32
3.35
3.14
3.19
3.21
3.15
3.17
3.13
3.03
2.95
3.16
3.06
3.27
3.28
2.96
3.01
2.96
3.03
2.99
3.00
2.84
2.87
2.90
2.95
3.08
3.13
2.75
2.83
2.78
2.84
2.73
2.82
2.60
2.67
2.68
2.77
2.86
2.94
2.97
3.03
2.89
3.00
2.87
2.90
2.69
2.77
2.88
2.89
2.97
3.07
3.30
3.27
3.20
3.14
3.16
3.14
2.95
2.97
3.23
3.05
3.25
3.28
2.58
2.69
2.53
2.63
2.46
2.58
2.24
2.48
2.57
2.54
2.57
2.70
2.66
2.79
2.67
2.76
2.56
2.67
2.54
2.59
2.54
2.61
2.69
2.84
2.73
2.79
2.71
2.77
2.64
2.71
2.63
2.67
2.67
2.71
2.69
2.79
2.93
2.98
2.92
2.95
2.87
2.90
2.73
2.76
2.89
2.86
2.97
3.04
Outcome expectations. “What will happen if you eat more fruit and vegetables”: (1–5)
I will be in better shape if I
eat more fruit and
vegetables
I will like myself better if I
eat more fruit and
vegetables
I will get more energy if I
eat more fruit and
vegetables
I will lose weight if I eat
more fruit and vegetables
I will be better looking if I
eat more fruit and
vegetables
If I eat more fruit and
vegetables, so will the rest
of my family
My family will be pleased
if I eat more fruit and
vegetables
Outcome expectations
scale average
3.73
3.70
3.79
3.75
3.68
3.60
3.65
3.51
3.74
3.68
3.76
3.75
3.72
3.58
3.77
3.65
3.59
3.44
3.48
3.25
3.62
3.53
3.80
3.70
4.13
4.02
4.17
3.99
3.98
3.84
3.88
3.70
4.15
3.92
4.13
4.03
3.42
3.42
3.53
3.46
3.36
3.33
3.36
3.25
3.50
3.31
3.45
3.52
3.37
3.39
3.52
3.41
3.40
3.40
3.43
3.32
3.51
3.39
3.45
3.45
2.68
2.72
2.95
2.86
2.61
2.66
2.58
2.72
2.78
2.80
2.87
2.76
3.44
3.29
3.49
3.25
3.32
3.18
3.36
3.25
3.39
3.26
3.44
3.18
3.50
3.44
3.60
3.48
3.42
3.35
3.39
3.28
3.53
3.41
3.55
3.48
179
Figure 1 shows the number of sent text messages during each of the five
weeks, separately, for boys and girls. As expected from the reviewed literature, girls
participated more actively than boys. The usual attrition in reporting during the course of
the study was observed for both girls and boys. The proportion of the possible participant
SMSs sent varied widely between schools, from 30 to 70%. This variation was unrelated
to experimental treatments. In the SMS group, individual schools sent from 30 to 70%;
and, in the SMS/education group, from 33 to 66% of the possible SMSs.
Figure 1
Number of sent text messages (in percent) per week, comparing boys and girls
Hierarchical regression analyses using the change in behavior (i.e., fruit and
vegetable intake), self-efficacy and outcome expectations from the before to the after
survey revealed that Hypothesis 1a, predicting a direct effects of the interventions, is not
supported by the data. When adding dummy variables for the intervention groups to a
model that only included the baseline level of the dependent variable, this led to a non180
significant F-change for behavior change (F-change = 0.752, 2 d.f., p = .472), change in
self-efficacy (F-change = 0.423, 2 d.f., p = .655), and change in outcome expectations (Fchange = 0.925, 2 d.f., p = .397). Further adding gender led to a significant F-change for
behavior change (F-change = 16.236, 1 d.f., p < .001) and in self-efficacy (F-change =
12.519, 1 d.f., p < .001), but not in outcome expectations (F-change = 0.917, 1 d.f., p =
.338). However, gender did not moderate the impacts of the intervention. When adding
the interactions between gender and the two dummy variables representing different
interventions to the regression model, this led to a non-significant F-change for behavior
change (F-change = 0.337, 2 d.f., p = .714), change in self-efficacy (F-change = 0.591, 2
d.f., p = .554), and change in outcome expectations (F-change = 0.720, 2 d.f., p = .487). It
also follows from these results that Hypothesis 1b, predicting that the combined
SMS/education treatment would lead to bigger changes in behavior, self-efficacy and
outcome expectations regarding fruit and vegetable intake than the SMS treatment alone,
is also not supported. The significant gender effect reflects that girls experienced a more
favorable change in behavior and self-efficacy than boys during the intervention. The
non-significant interaction effects show that this effect was independent on experimental
conditions. Hence, it appears to be a variant of the “Hawthorne” effect (McCarney et al.,
2007).
The first multiple regression analysis reported in Table 3 reveals that,
consistent with Hypothesis 2, participants sending more than half of the possible text
messages, but only these, significantly increased their intake of fruit and vegetables from
the pre to the post measurement (p < .01). Further, and unexpectedly, participants sending
between 10 and 50% of the text messages experienced a significant drop in self-efficacy
in relation to fruit and vegetable intake (p < .05). Also, participants sending less than 10%
of the possible text messages reported a significant drop in outcome expectations in
181
relation to fruit and vegetable intake (p < .05). With a significant improvement in
behavior for those sending more than 50% of the possible text messages, but no
improvement in self-efficacy and outcome expectations, Hypothesis 2 is only partly
supported. Finally, Table 3 reveals the expected regression towards the mean: The higher
the level of behavior, self-efficacy and outcome expectations at baseline, the less they
improved from the pre to the post survey.
Table 3
The impact of number of sent text messages by participants in the two intervention groups on changes in
reported fruit and vegetable intake, self-efficacy and outcome expectations from the pre (t1) to the post (t2)
measurement while controlling for the baseline level of the dependent variable (t 1)
B
Std. Error
Beta
t
sig.
Change in fruit and vegetable intake
.19
(Constant)
-0.11
0.04
1
-0.08
0.08
-0.06
Low
2
Medium
High3
Behavior(t1)
-2.503
.012
-0.03
-0.928
.354
0.08
-0.02
-0.754
.451
0.17
0.06
0.08
2.699
.007
-0.43
0.03
-0.43
-16.406
.000
Change in self-efficacy
(Constant)
.21
0.07
0.03
Low
-0.11
0.07
Medium
-0.13
High
Self-efficacy(t1)
2.152
.032
-0.04
-1.584
.113
0.06
-0.06
-2.110
.035
0.02
0.05
0.01
0.314
.753
-0.37
0.02
-0.46
-17.617
.000
Change in outcome expectations
.20
(Constant)
-0.06
0.03
Low
-0.17
0.07
Medium
-0.05
High
Outcome Expectations(t1)
R2 adj.
-1.643
.101
-0.07
-2.466
.014
0.06
-0.03
-0.880
.379
0.00
0.05
0.00
-0.036
.971
-0.36
0.02
-0.45
-16.872
.000
1
Less than 10% (0–3 text messages). 2 Between 10 and 50% (4–17 text messages). 3 Over 50% (18 to 35
text messages).
182
Despite the substantial gender difference in the level of activity in the
intervention (Figure 1), gender did not moderate the relationships reported in Table 3.
When adding the interactions between gender and the three dummy variables representing
different levels of SMS activity to the regression model in Table 3, extended with gender,
this led to a non-significant F-change for behavior change (F-change = 0.416, 3 d.f., p =
.742), change in self-efficacy (F-change = 1.081, 3 d.f., p = .356), and change in outcome
expectations (F-change = 1.026, 3 d.f., p = .380). Hence, Hypothesis 3 is only partly
confirmed by the data.4
7.5. Discussion
This study has investigated the potential of text messaging as a two-way
communication tool in a feedback intervention, engaging participants in a learning-bydoing process to facilitate changes in health behavior. The study also investigated the
impacts of the intervention on two presumed mediators of intervention effects on
behavior: self-efficacy and outcome expectations. Finally, the study provides unique
insight, in the context of a field test, into how the impacts of an intervention depend on
participants’ activity level in the intervention.
Prior research has used tailored feedback via text messaging in clinical
settings with adults (see for instance de Jongh, Gurol-Urganci, Vodopivec-Jamsek, Car,
4
Since the data covered two different age groups (5th graders and 9th graders), it is also possible to
test for the importance of age differences. However, a similar hierarchical regression analysis
found no significant interactions between age and the text messaging activity level. These
statistical analyses can be acquired from the first author.
183
& Atun, 2012; Krishna, Boren, & Balas, 2009, pp., for a variety of behaviours; Nobis et
al., 2013; Nundy, Dick, Solomon, & Peek, 2013), but less in interventions targeting
adolescents (though see Lana et al. (2013) on cancer prevention, Haug et al. (2009) on
smoking cessation and Woolford et al. (2010) and Bauer et al. (2010) on treatment of
obesity). This study tests a community-based intervention promoting healthy eating and
thereby preventing (rather than treating) overweight and obesity (see O'Dea, 2005, for a
distinction). The main finding is that a feedback intervention using text messaging can
increase adolescents’ fruit and vegetable intake, but only if participants are sufficiently
active; in this case reflected in them sending more than half of the possible text messages.
The found effect sizes are small, but worthwhile, as also found in previous studies using
feedback via text messaging to promote fruit and vegetable intake among adolescents
(Bech-Larsen & Grønhøj, 2013) and in studies using Internet-delivered feedback
interventions (Gorely, Nevill, Morris, Stensel, & Nevill, 2009; Hamel & Robbins, 2013).
Similar to Bech-Larsen and Grønhøj (2013), this study also found that the increase in
adolescents’ consumption of fruit and vegetables during the intervention was lower the
higher the pre-intervention intake, reflecting that there is a limit to how much fruit and
vegetables one can eat during a day (statistically, a regression towards the mean).
Substantially, this means that those most in need of change actually changed the most.
Further, since most adolescents do not eat the recommended amount of fruit and
vegetables, there is plenty of room for improvement and a need for interventions that
close the gap.
The provided information about benefits of eating more fruits and
vegetables, especially in the education session by a dietician, addressed participants’
outcome expectations directly, and giving feedback is a recognized means to improve
self-efficacy. However, contrary to expectations, the identified behavior change among
184
the most active participants was not mediated through changes in self-efficacy or outcome
expectations. The only significant change in self-efficacy produced by the intervention
was a negative impact among participants with a medium level of activity (i.e., sending
between 10 and 50% of the text messages). One might speculate that the loss in selfefficacy was due to these participants experiencing the daily text messages as unpleasant
reminders about their failure to perform a challenging task. When reminded of one’s lack
of previous goal achievement (in this case, regarding sending SMS reports), one might
develop a lack of confidence in one’s ability to succeed with this activity, resulting in a
negative cycle of more omissions, lower self-efficacy, and perhaps a feeling of
helplessness (Abramson, Seligman, & Teasdale, 1978; Maier & Seligman, 1976). This is
a serious “boomerang effect” (Ringold, 2002) and a risk that should be considered
thoroughly when planning interventions and which should be investigated further in
targeted research. This study cannot determine whether participants experiencing a drop
in self-efficacy are also more prone to develop personal inefficacy (Bandura, 1982) in
other behavior change experiences, but the risk of such negative “spillovers” from health
interventions (Bech-Larsen & Kazbare, 2014; O'Dea, 2005) is also worth exploring in
future studies.
Also contrary to expectations, the only change in outcome expectations was
a significant drop for the least active participants (those sending less than 10% of the text
messages). It seems likely that this boomerang effect among pupils who never really got
started with the intervention reflects denial. Denial is a psychological defense mechanism,
a coping strategy to protect one’s self-esteem that arises to avoid the feeling of failure, in
this case, because of facing the fact that one is probably not able to make the necessary
changes to obtain the self-set goal (Baumeister, Dale, & Sommer, 1998). Specifically, the
denier adjusts his/her outcome expectations with doubts that participation will make a
185
difference, thus justifying not participating actively in the intervention (Abelson, 1959).
Obviously, also this unexpected consequence of failing to complete the intervention needs
to be researched more in future studies.
One might speculate that negative results like the ones just described could
be avoided in future intervention studies through supplementary support measures that
assure sufficient coping with the requests of the intervention. In the present case, an
activity level of 50% (i.e., submitting half of the possible text messages) was sufficient to
ensure small, but positive changes in diets.
The analyses revealed significant differences between genders in changes in
behavior and self-efficacy between the two surveys and also in level of activity in the
interventions. However, there were no differences in intervention impacts. Hence, a
“Hawthorn effect” for girls, where participants respond to being part of a study, rather
than to the specific interventions (Wickstrom & Bendix, 2000), cannot be ruled out.
It is a general challenge for intervention studies to keep participants
motivated and make sure that they are sufficiently engaged to ensure positive behavior
change and avoid possible boomerang effects. Response rates in text messaging studies
vary for a lot of reasons, such as context and whether it is one- or two-way
communication (Whitford et al., 2012). This study required one reminder-report-feedback
interaction with each participant per day; which, according to Bosworth and Shaw (2012),
is appropriate to motivate people without generating a too-large burden for them.
Participants also had breaks from reporting every other week during the 11-week
intervention period in order to avoid feeling too monitored.
Future campaign planners should pay particular attention to gender
difference in the activity level which in this study dropped, from week 1 to 5, from 76%
186
to 41% for boys and from 91% to 60% for girls (see Figure 1). The drop in activity level
can be a reactance effect (Brehm & Brehm, 1981; Dowd, 2002), if participants feel that
reporting on their fruit and vegetable intake is threatening their freedom of choice or of
not having to pay attention to food intake during the day. In order to reduce the risk of
reactance, the feedback was designed to not include any socially pressuring information,
such as comparing reported intake to recommended intake or to the average intake of
others, but only intake relatively to a privately set goal. The only additional measure
taken to keep participants active in the intervention was a prize to the class that sent in the
most SMSs during the intervention, but no information was collected to test the
effectiveness of the prize. Hence, it might be useful to test different approaches to counter
attrition in similar interventions in the future, including different approaches targeting
boys and girls.
Out of 233 participants in the intervention groups who did not send one
single text message, 154 did not register the first week and 79 registered, but did not send
any text messages. The reason for this lack of activity can be participants’ unwillingness
and reactance, but technical problems with registration and reporting can also not be ruled
out. If a report were not accepted by the computer system due to deviations from the
standardized answer template, participants would receive a standardized error message
encouraging them to try again and repeating the intended template. This might not have
been enough, even though participants also received a pamphlet with questions and
answers and had access to a hotline. Future studies might use qualitative assessments
from participants with different activity levels and/or quantitative measurements of
participants’ motivation to engage in behavioral change as input to develop strategies for
keeping participants active in the intervention and motivate them to be highly active.
187
Special attention should be paid to differences between genders, since boys were on
average less active than girls from the outset.
This study found no significant effects of a 45-minute nutritional education
session over and above participating in the text messaging intervention. The lack of effect
of nutritional education of this type is consistent with previous studies, including some
focusing on education targeting fruit and vegetable intake through text messaging (Head,
Noar, Iannarino, & Grant Harrington, 2013). This suggests that it is not necessary to add
education from a dietician in order to obtain the desired effects of a text messaging
intervention, at least not in the studied context. One might speculate that the reason might
be that these adolescents already had sufficient knowledge beforehand about the health
benefits of eating the recommended amount of fruit and vegetables.
Like all other studies, this one has its limitations. First of all, the reported
intervention is relatively short and conducted in only one country. Even though it is
recommended to keep interventions targeting adolescents relatively short (less than 16
weeks in order to obtain significant positive effects, according to Stice et al. (2006)), it
could have been informative to include a long-term follow up survey in the study design.
It has been suggested that using a competition to incentivize, in this case, a higher activity
level might increase socially desirable responses (Raju, Rajagopal, & Gilbride, 2010).
However, when the prize depends upon collective rather than individual responses and
sending text messages rather than on increased intake of fruit and vegetables, such
adverse effects seem less likely. Self-reported behavior (as well as self-efficacy and
outcome expectations) is fallible since respondents might neither be able nor willing to
accurately report their own behavior (Thøgersen, 2008). It was attempted to reduce this
problem by using multiple items and asking about the frequency of eating fruit and
188
vegetables at average meals and snacks which is easier for adolescents (and adults) to
report than, for example, portion sizes (Lechner, Brug, & De Vries, 1997; Mullarkey,
Johnson, & Hackett, 2007).
Since schools were the unit of randomization, one can imagine that there are
idiosyncratic school-level effects. However, the number of schools (17, leading to 5 or 6
schools in each condition) is too small to provide sufficient statistical power to separate
school effects from intervention effects in a multilevel model. Also, given that there are
5–6 schools in each condition, it seems extremely unlikely that idiosyncratic school
effects should produce noticeable biases in the estimations of the intervention effects.
Even though text messaging is a popular communication form, its role in
health promoting services might be outpaced by apps in smartphones which can provide
immediate tailored feedback in a more aesthetic and detailed manner than text messages.
A recent review of the use of smartphone technology in promoting physical activity found
modest effects due to small sample sizes, short-term studies and a lack of randomized
controlled trials (Bort-Roig, Gilson, Puig-Ribera, Contreras, & Trost, 2014). One can
easily get carried away with new technology’s promising features. However, it is
important to design studies that encompass the theoretical underpinnings of behavior
change and the role of immediate feedback and, at the same time, take advantage of the
ease and accessibility of new technology. Ensuring easy practical implementation and low
costs, two-way communication outlets provide promising means for health promotion in
the future.
In conclusion, this study has provided new insights on the usefulness of text
messaging for health promotion targeting adolescents and especially on the importance of
189
participants’ activity levels for self-efficacy, outcome expectations and, not least, positive
behavioral changes.
Funding
This work was supported by the Danish Ministry of Science, Technology
and Innovation, grant no. 09/061357, as part of the research project “Step by step changes
of children’s preferences towards healthier food.”
Acknowledgements
The authors would like to thank the participants and schools. Special thanks
to project leader Tino Bech-Larsen for access to the data and useful comments on an
earlier draft, to student assistants Jacob Heiss Rosendahl, Astrid Refsgaard, Ken
Jørgensen and Christina Bæklund for their work in relation to the data collection and to
Birgitte Steffensen for proofreading the manuscript.
190
References
Abelson, R. P. (1959). Models of resolution of belief dilemmas. Journal of Conflict
Resolution, 3, 343-352.
Abramson, L. Y., Seligman, M. E., & Teasdale, J. D. (1978). Learned helplessness in
humans: Critique and reformulation. Journal of Abnormal Psychology, 87(1), 4974.
Bandura, Albert. (1977). Social learning theory (1st ed.). Englewood Cliffs, NJ: PrenticeHall.
Bandura, Albert. (1982). Self-efficacy mechanism in human agency. American
Psychologist, 37(2), 122-147.
Bandura, Albert. (1986). Social foundations of thought and action: A social cognitive
perspective. Englewood Cliffs, NJ: Prentice-Hall.
Bandura, Albert. (2004). Health promotion by social cognitive means. Health Education
& Behavior 31(2), 143-164.
Baranowski, T., Lin, L. S., Wetter, D. W., Resnicow, K., & Hearn, M. D. (1997). Theory
as mediating variables: Why aren't community interventions working as desired?
Annals of Epidemiology, 7(7, Supplement), 89-95.
Bauer, S, de Niet, J, Timman, R, & Kordy, H. (2010). Enhancement of care through selfmonitoring and tailored feedback via text messaging and their use in the treatment
of childhood overweight. Patient Education and Counseling, 79, 313-319.
Baumeister, R. F., Dale, K., & Sommer, K. L. (1998). Freudian defense mechanisms and
empirical findings in modern social psychology: Reaction formation, projection,
displacement, undoing, isolation, sublimation and denial. Journal of Personality,
66(6), 1081.
Bazzano, L. A. (2006). The high cost of not consuming fruits and vegetables. Journal of
the American Dietetic Association, 106(9), 1364-1368.
Bech-Larsen, Tino, & Grønhøj, Alice. (2013). Promoting healthy eating to children: A
text message (SMS) feedback approach. International Journal of Consumer
Studies, 37(3), 250-256. doi: 10.1111/j.1470-6431.2012.01133.x
Bech-Larsen, Tino, & Kazbare, Laura. (2014). Spillover of diet changes on intentions to
approach healthy food and avoid unhealthy food. Health Education, 114(5), 367377. doi: 10.1108/he-04-2013-0014
191
Bere, Elling, & Klepp, Knut-Inge. (2004). Correlates of fruit and vegetable intake among
Norwegian schoolchildren: Parental and self-reports. Public Health Nutrition,
7(8), 991-998. doi: 10.1079/phn2004619
Birnbaum, A., Lytle, L., Story, M., Perry, C., & Murray, D. (2002). Are differences in
exposure to a multicomponent school-based intervention associated with varying
dietary outcomes in adolescents? Health Education & Behavior 29(4), 427-443.
Bort-Roig, Judit, Gilson, Nicholas D, Puig-Ribera, Anna, Contreras, Ruth S, & Trost,
Stewart G. (2014). Measuring and influencing physical activity with smartphone
technology: A systematic review. Sports Medicine, 44(5), 671-686. doi:
10.1007/s40279-014-0142-5
Bosworth, H., & Shaw, R. (2012). Short message service (SMS) text messaging as an
intervention medium for weight loss: A literature review. Health Informatics
Journal, 18(4), 235-250.
Brehm, S. S., & Brehm, J. W. (1981). Psychological reactance: A theory of freedom and
control. San Diego: Academic Press.
Brislin, Richard W. (1970). Back-translation for cross-cultural research. Journal of CrossCultural Psychology, 1(3), 185-216. doi: 10.1177/135910457000100301
Cerin, Ester, Barnett, Anthony, & Baranowski, Tom. (2009). Testing theories of dietary
behavior change in youth using the mediating variable model with intervention
programs. Journal of Nutrition Education and Behavior, 41(5), 309-318. doi:
10.1016/j.jneb.2009.03.129
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple
regression/correlation analysis for the behavioral sciences (3rd ed.). Mahwah,
NJ: Lawrence Erlbaum.
Contento, Isobel R., Randell, Jill S., & Basch, Charles E. (2002). Review and analysis of
evaluation measures used in nutrition education intervention research. Journal of
Nutrition Education and Behavior, 34(1), 2-25. doi: 10.1016/S14994046(06)60220-0
Davie, Ronald, Panting, Charlotte, & Charlton, Tony. (2004). Mobile phone ownership
and usage among pre-adolescents. Telematics and Informatics, 21(4), 359-373.
doi: 10.1016/j.tele.2004.04.001
de Jongh, Thyra, Gurol-Urganci, Ipek, Vodopivec-Jamsek, Vlasta, Car, Josip, & Atun,
Rifat. (2012). Mobile phone messaging for facilitating self-management of longterm illnesses. The Cochrane database of systematic reviews, 12, CD007459.
Dowd, E. T. (2002). Psychological reactance in health education and promotion. Health
Education Journal, 61(2), 113-124.
192
Faulkner, Xristine, & Culwin, Fintan. (2005). When fingers do the talking: A study of
text messaging. Interacting with Computers, 17(2), 167-185. doi:
10.1016/j.intcom.2004.11.002
Fitzgerald, A., Heary, C., Kelly, C., Nixon, E., & Shevlin, M. (2013). Self-efficacy for
healthy eating and peer support for unhealthy eating are associated with
adolescents’ food intake patterns. Appetite, 63, 48-58. doi:
10.1016/j.appet.2012.12.011
Fjeldsoe, Brianna S., Marshall, Alison L., & Miller, Yvette D. (2009). Behavior change
interventions delivered by mobile telephone short-message service. American
Journal of Preventive Medicine, 36, 165-173.
France, Alan. (2004). Young people. In S. Fraser, V. Lewis, S. Ding, M. Kellett & C.
Robinson (Eds.), Doing research with children and young people (pp. 175-190).
London: SAGE Publications.
Geller, K., & Dzewaltowski, D. (2010). Examining elementary school-aged children's
self-efficacy and proxy efficacy for fruit and vegetable consumption. Health
Education & Behavior, 37(4), 465-478. doi: 10.1177/1090198109347067
Gorely, Trish, Nevill, Mary E, Morris, John G, Stensel, David J, & Nevill, Alan. (2009).
Effect of a school-based intervention to promote healthy lifestyles in 7-11 year old
children. The International Journal of Behavioral Nutrition and Physical Activity,
6(1), 5-16. doi: 10.1186/1479-5868-6-5
Hamel, L. M., & Robbins, L. B. (2013). Computer- and web-based interventions to
promote healthy eating among children and adolescents: A systematic review.
Journal of Advanced Nursing, 69(1), 16-30.
Haug, Severin, Meyer, Christian, Schorr, Gudrun, Bauer, S, & John, Ulrich. (2009).
Continuous individual support of smoking cessation using text messaging: A pilot
experimental study. Nicotine & tobacco research: Official Journal of the Society
for Research on Nicotine and Tobacco, 11(8), 915-923.
Head, Katharine J, Noar, Seth M, Iannarino, Nicholas T, & Grant Harrington, Nancy.
(2013). Efficacy of text messaging-based interventions for health promotion: A
meta-analysis. Social Science & Medicine (1982), 97, 41-48.
Kluger, A.N., & DeNisi, A. (1996). The effects of feedback interventions on
performance: A historical review, a meta-analysis and a preliminary feedback
intervention theory. Psychological Bulletin, 119(2), 254-284.
Krishna, Santosh, Boren, Suzanne Austin, & Balas, E Andrew. (2009). Healthcare via cell
phones: A systematic review. Telemedicine journal and e-health: The official
journal of the American Telemedicine Association, 15(3), 231-240. doi:
10.1089/tmj.2008.0099
193
Lana, Alberto, del Valle, Maria Olivo, Lopez, Santiago, Faya-Ornia, Goretti, & Lopez,
Maria Luisa. (2013). Study protocol of a randomized controlled trial to improve
cancer prevention behaviors in adolescents and adults using a web-based
intervention supplemented with SMS. BMC Public Health, 13(1), 357.
Lechner, L., Brug, J., & De Vries, H. (1997). Misconceptions of fruit and vegetable
consumption: Differences between objective and subjective estimation of intake.
Journal of Nutrition Education, 29(6), 313-320.
Locke, Edwin A, Shaw, Karyll N, Saari, Lise M, & Latham, Gary P. (1981). Goal setting
and task performance: 1969-1980. Psychological bulletin, 90(1), 125-152. doi:
10.1037/0033-2909.90.1.125
Luszczynska, A., Tryburcy, M., & Schwarzer, R. (2007). Improving fruit and vegetable
consumption: A self-efficacy intervention compared to a combined self-efficacy
and planning intervention. Health Education Research, 22, 630-638.
Lynch, Christel, Kristjansdottir, Asa Gudrun, te Velde, Saskia J, Lien, Nanna, Roos, Eva,
Thorsdottir, Inga, . . . Yngve, Agneta. (2014). Fruit and vegetable consumption in
a sample of 11-year-old children in ten European countries – the PRO GREENS
cross-sectional survey. Public Health Nutrition, 17(11), 2436-2444. doi:
10.1017/S1368980014001347
Maier, S. F., & Seligman, M. E. P. (1976). Learned helplessness: Theory and evidence.
Journal of Experimental Psychology, 105(1), 3 - 46.
McCarney, R., Warner, J., Iliffe, S., van Haselen, R., Griffin, M., & Fisher, P. (2007).
The Hawthorne effect: A randomised, controlled trial. BMC Medical Research
Methodology, 7, 30-38.
Mullarkey, D., Johnson, B., & Hackett, A. (2007). Portion size selection of fruits and
vegetables by 9- to 10-year-old children in Liverpool. Journal of Human Nutrition
& Dietetics, 20(5), 459-466.
Nobis, Stephanie, Lehr, Dirk, Ebert, David Daniel, Berking, Matthias, Heber, Elena,
Baumeister, Harald, . . . Riper, Heleen. (2013). Efficacy and cost-effectiveness of
a web-based intervention with mobile phone support to treat depressive symptoms
in adults with diabetes mellitus type 1 and type 2: Design of a randomised
controlled trial. BMC psychiatry, 13(1), 306-306. doi: 10.1186/1471-244x-13-306
Nundy, Shantanu, Dick, Jonathan J, Solomon, Marla C, & Peek, Monica E. (2013).
Developing a behavioral model for mobile phone-based diabetes interventions.
Patient Education and Counseling, 90(1), 125.
O'Dea, J. A. (2005). Prevention of child obesity: 'First, do no harm'. Health education
research, 20(2), 259-265.
194
Phau, I., & Teah, M. . (2009). Young consumers’ motives for using SMS and perceptions
towards SMS advertising. Direct Marketing, 3(2), 97-108.
Preston, Karen E, Walhart, Tara A, & O'Sullivan, Ann L. (2011). Prompting healthy
behavior via text messaging in adolescents and young adults. American Journal of
Lifestyle Medicine, 5(3), 247-252. doi: 10.1177/1559827610392325
Raju, S., Rajagopal, P., & Gilbride, T. (2010). Marketing healthful eating to children: The
effectiveness of incentives, pledges and competitions. Journal of Marketing, 74,
93-106.
Rasmussen, Mette, Krolner, Rikke, Klepp, Knut-Inge, Lytle, L, Brug, Johannes, Bere, E,
& Due, Pernille. (2006). Determinants of fruit and vegetable consumption among
children and adolescents: A review of the literature. Part I: quantitative studies. Int
J Behav Nutr Phys Act, 3, 22.
Resnicow, K., Davis-Hearn, M., Smith, M., Baranowski, T., Lin, L. S., Baranowski, J., . .
. Wang, D. T. (1997). Social-cognitive predictors of fruit and vegetable intake in
children. Health Psychology, 16(3), 272-276.
Ringold, D. J. (2002). Boomerang effect: In response to public health interventions: Some
unintended consequences in the alcoholic beverage market. Journal of Consumer
Policy, 25, 27-63.
Schultz, P. Wesley. (1999). Changing behavior with normative feedback interventions: A
field experiment of curbside recycling. Basic and Applied Social Psychology, 21,
25-36.
Shapiro, Jennifer R., Bauer, S., Hamer, Robert M., Kordy, Hans, Ward, Dianne, & Bulik,
Cynthia M. (2008). Use of text messaging for monitoring sugar-sweetened
beverages, physical activity and screen time in children: A pilot study. Journal of
Nutrition Education and Behavior, 40(6), 385-391. doi:
10.1016/j.jneb.2007.09.014
Statistik, Danmarks. (2011). Befolkningens brug af internet 2010 [The population's use of
Internet 2010]. Retrieved February 5, 2015, from
http://www.dst.dk/da/Statistik/Publikationer/VisPub.aspx?cid=018686
Stice, Eric, Shaw, Heather, & Marti, C. Nathan. (2006). A meta-analytic review of
obesity prevention programs for children and adolescents: The skinny on
interventions that work. Psychological Bulletin, 132(5), 667-692.
Story, M., Mays, R. W., Bishop, D. B., Perry, C. L., Taylor, G., Smyth, M., & Gray, C.
(2000). 5-a-day power plus: Process evaluation of a multicomponent elementary
school program to increase fruit and vegetable consumption. Health Education &
Behavior, 27(2), 187-200.
195
Sundhedsstyrelsen. (2008). Undersøgelse af 11-15 åriges livsstil og sundhedsvaner 19972006 [A study of 11-15 year-olds’ life styles and health habits 1997-2006]
Copenhagen Sundhedsstyrelsen [Danish Health and Medicines Authority].
Thøgersen, John. (2008). Social norms and cooperation in real-life social dilemmas.
Journal of Economic Psychology, 29(4), 458-472. doi: 10.1016/j.joep.2007.12.004
Thøgersen, John, & Grønhøj, Alice. (2010). Electricity saving in households - A social
cognitive approach. Energy Policy, 38, 7732-7743.
Watten, R. G., Kleiven, J., Fostervold, K. I., Fauske, H., & Volden, F. (2008). Gender
profiles of internet and mobile phone use among Norwegian adolescents.
Seminar.Net: Media, Technology & Life-Long Learning, 4(3), 1-9.
Whitford, H. M., Donnan, P. T., Symon, A. G., Kellett, G., Monteith-Hodge, E.,
Rauchhaus, P., & Wyatt, J. C. (2012). Evaluating the reliability, validity,
acceptability and practicality of SMS text messaging as a tool to collect research
data: Results from the Feeding Your Baby project. Journal of the American
Medical Informatics Association, 19(5), 744-749.
WHO. (2003). Diet, nutrition and the prevention of chronic diseases. Geneva: WHO
Wickstrom, G., & Bendix, T. (2000). The "Hawthorne effect" - what did the original
Hawthorne studies actually show? Scandinavian Journal of Work, Environment &
Health, 26(4), 363-367.
Woolford, Susan J, Clark, Sarah J, Strecher, Victor J, & Resnicow, Kenneth. (2010).
Tailored mobile phone text messages as an adjunct to obesity treatment for
adolescents. Journal of Telemedicine and Telecare, 16(8), 458-461. doi:
10.1258/jtt.2010.100207
Young, E., Fors, S. W., Fasha, E., & Hayes, D. (2004). Associations between perceived
parent behaviors and middle school student fruit and vegetable consumption.
Journal of Nutrition Education and Behavior (36), 2-12.
196
Appendix A
Counting form handed out to participants in the SMS intervention
1 enhed = 1 unit = 100 grams
197
198
8. CONCLUSION AND IMPLICATIONS
In order to respond to the soaring levels of overweight and obesity among
adolescents, it is important to understand adolescent healthy eating by accounting for
influencing factors, barriers and facilitators. Such an understanding is a necessary
foundation for interventions to improve adolescents’ healthy eating behaviour. Therefore,
this PhD thesis has contributed with knowledge regarding the socialisation of healthy
eating behaviours and the social influences and conditions for behaviour change in
relation to adolescents’ healthy eating, more specifically fruit and vegetable intake.
Past consumer socialisation research has used a cognitive development
approach to determine what is learned at particular developmental stages (John, 1999),
while social learning research has focused on modelling to explain how socialisation
takes place (Moschis & Churchill, 1978). Parents were identified as primary socialisation
agents and friends, school, media as secondary socialisation agents in these approaches.
Hence, previous research has tended to study the development of healthy eating
behaviour and the social influences and conditions for behaviour change as aspects
somewhat isolated from each other. This thesis contributes with bringing these aspects of
the healthy eating context closer together.
Overall, this thesis deals with the socialisation of healthy eating behaviour,
how it is influenced and how it can be improved. This was investigated in the context of
the research project “Step by step changes of children’s preferences towards healthier
food” where a mixed method approach provided qualitative data (semi-structured
interviews with families and a practical exercise) and quantitative data (a feedback
intervention involving almost 1500 adolescents and pre- and post-intervention surveys).
The findings are presented in three research papers. The first research paper explored the
199
consumer socialisation of healthy eating within the family and especially the adolescents’
and parents’ roles in this regard (paying attention to environmental factors such as food
culture and social context). The second research paper investigated the social influence of
parents and friends and compared it to the influence of the adolescents’ own self-efficacy
and outcome expectations in order to obtain a more comprehensive insight into what
affects adolescents’ healthy eating. The third research paper investigated the effects of a
feedback intervention employing text messaging on adolescents’ behaviour, self-efficacy
and outcome expectations regarding healthy eating. Hence, jointly the papers contribute
with insights into how adolescents’ more or less healthy eating is formed and can be
influenced.
The most important contribution of this thesis is the increased knowledge
and insights into adolescent healthy eating, and more specifically their fruit and vegetable
intake, which comes about when linking consumer socialisation research and behaviour
change theories in the framework of social cognitive theory. This approach has enabled a
more holistic view on healthy eating by emphasizing that human agency operates within
an interdependent causal structure. Adolescents’ eating behaviour is therefore not studied
in isolation, but in its social and cultural context, alongside the adolescents’ personal
motivation, with the aim to understand how behaviour develop and can be influenced.
The common assumption in consumer socialisation research that adolescents are
increasingly influenced by peers at the expense of parents is not confirmed by this
research. At least it seems to not result in the former becoming dominating in the context
of healthy eating. Parents remain the main influencer of healthy eating during
adolescence and, hence, an intra-familiar systemic approach seems the most useful when
studying healthy eating socialisation. It was also found that especially what parents do
(perceived descriptive norms) has a significant impact on adolescents’ healthy eating
200
behaviour, along with the adolescent’s self-efficacy and outcome expectations. Previous
studies of consumer socialisation have mainly looked into the cognitive abilities and not
so much the adolescents’ own motivations in relation to the socialisation process and
possible behaviour changes. Hence, the socialisation process appears more automatic than
it probably is.
Furthermore, it was found that when participants engage actively in a
feedback intervention using text messaging, such an intervention can influence
adolescents’ fruit and vegetable intake significantly. As a physical action – putting
healthy foods in your mouth – healthy eating is apparently not a terribly complex
behaviour. Yet, with a more holistic view – as in this thesis – trying to take personal
factors and environmental factors such as social structures and social influences into
account, it appears that healthy eating is actually more complex than it may seem. This
should not be underestimated when trying to fight the rising levels of overweight. In the
thesis it was found that personal factors such as self-efficacy (the belief in one’s own
abilities to change behaviour) are very important for healthy eating, but also the social
influence of parents (especially perceived descriptive norms). It is important that health
promotors and health policy makers are aware of such findings and of the barriers and
facilitators of healthy eating in general.
8.1. Specific research paper contributions
The first research paper, in Chapter 5, explores adolescents’ and parents’
roles in healthy eating socialisation within the family based on participation in an
intervention three months earlier. By assessing both the adolescent’s and the parents’
view on their roles in the family regarding healthy eating, the research paper provides a
201
thorough – and rare – empirical account of consumer socialisation. A few previous
studies of healthy eating have described the socialisation processes as bi-directional (e.g.
Kuczynski & Parkin, 2006) but only very rarely provided an in-depth analysis of what
this actually means and how it takes place. Hence, this research paper investigates
adolescents’ and parents’ accounts of awareness and involvement in healthy eating and
shows that adolescents and parents identified several causes of awareness and
involvement in healthy eating: new or re-activated health knowledge, visualisation of
amounts, self-regulation and planning. In line with Nørgaard et al.’s (2007) study on food
buying, results show that adolescents adopted two roles in terms of family socialisation: a
direct influencer role placing demands on parents or a cooperative role helping the
parents. Parents initiated dialogues with family members about healthy eating and felt
responsible as role models often fulfilling the adolescent’s demands and acknowledging
their help. Parents also regulated family members’ eating behaviour through shopping,
cooking and discussions about healthy eating with family members. The study finds that
parents still have the upper hand, when it comes to healthy eating, but with adolescents as
active players. The main contribution of Chapter 5 is the utilisation of the adolescent’s
immediate family as a unit of analysis when explaining in detail how socialisation efforts
materialise within the family in the form of strategies and roles. These findings have
practical implications for health promoters who should take into account the family
support (or lack hereof), when designing interventions and evaluating the outcomes.
The second research paper, in Chapter 6, offers two main contributions: The
first is the finding that parents continue to be the main influencer on adolescents’ healthy
eating (both for the younger and older adolescents) with what they do being more
important than what they say. Thereby, this study rejects a well-known assumption of
consumer socialisation research – namely that parents are the main influencers in
202
childhood with friends gradually taking over and becoming dominant in adolescence
(Gitelson & McDermott, 2006; John, 1999) – when it comes to healthy eating. Previous
studies have underlined the importance of social modelling and social influence on food
intake (e.g. Cruwys, Bevelander, & Hermans, 2015; Higgs, 2015), but have treated the
social influence from parents and friends separately, with a few exceptions (see for
instance Salvy, Elmo, Nitecki, Kluczynski, & Roemmich, 2011). This study contributes
with an analysis of the influence of “entangled social spheres” on adolescents’ healthy
eating. The findings suggest parents’ awareness of their impact on adolescent healthy
eating should be reinforced and that through targeting parents’ behaviour, health
promoters might also impact adolescents’ behaviour in a positive direction.
The other important contribution of this study is the inclusion of
adolescents’ self-efficacy and outcome expectations as important determinants for healthy
eating behaviour. By adapting Bandura’s (1986) triadic model of human agency, the
analysis includes the adolescents’ personal factors (self-efficacy and outcome
expectations), environment (e.g. social context for eating, food culture and perceived
social influence of parents and friends) and behaviour. Previous studies applying social
cognitive theory have found especially self-efficacy, but also outcome expectations
important for improving fruit and vegetable intake (e.g. Fitzgerald, Heary, Kelly, Nixon,
& Shevlin, 2013; Rasmussen et al., 2006; Resnicow et al., 1997). However, these studies
have not included the environmental context and social influence. Hence, it is an
important contribution to the literature of this study that it adds new knowledge about the
joint roles of important predictors of healthy eating among adolescents.
This study views healthy eating more holistically than previous studies since
no previous studies have included all the described factors in the same analysis. The two
main findings have practical implications for health promoters in order to optimise
203
intervention efforts and should be considered when designing interventions targeting
children and adolescents’ healthy eating. The study shows that it is important to build up
self-efficacy and favourable outcome expectations while also including parents as
socialisation agents and positive models as a means to increase healthy eating among
adolescents.
The third research paper, in Chapter 7, investigates the effects of a feedback
intervention employing text messaging on adolescents’ behaviour, self-efficacy and
outcome expectations regarding fruit and vegetable intake. Studies have found that SCT
serves well as a theoretical framework for behaviour change interventions (e.g. Contento,
Randell, & Basch, 2002; Michie, Johnston, Francis, Hardeman, & Eccles, 2008).
However, only few of the many SCT-based intervention studies apply feedback as a way
to enhance self-efficacy (see for instance Bech-Larsen & Grønhøj, 2013; Oenema &
Brug, 2003). This study found no significant direct effects of the intervention. However,
significant effects of the level of activity in the intervention – reflected in the number of
sent text messages – on intervention outcomes were found. Participants sending more
than half of the possible text messages significantly increased their fruit and vegetable
intake. However, contrary to expectations, the identified behaviour change was not
mediated through changes in self-efficacy or outcome expectations. It was also found,
unexpectedly, that participants sending between 10 and 50% of the possible text messages
experienced a significant drop in self-efficacy, and those sending less than 10%
experienced a significant drop in outcome expectations.
The first finding suggests that participants’ active engagement in an
intervention is crucial for its success. This study is one of the first to investigate, in the
context of a field test, how the impacts of an intervention depend on participants’ activity
level in the intervention. However, the negative results for self-efficacy and outcome
204
expectations for those least active in the intervention show that there is a risk of
boomerang effects when employing such interventions. Boomerang effects such as these
might be avoidable, but the findings show that future intervention studies definitely need
to develop supplementary support measures to assure sufficient coping with the requests
of the intervention.
Furthermore, this study found no significant effects of the 45-minute
nutritional education session over and above participating in the text messaging
intervention. This suggests that it is not necessary to add education from a dietician in
order to obtain the desired effects of a text messaging intervention, at least not in the
studied context. In the short description of the intervention leading up to the interviews in
Chapter 5, all three intervention groups had two visits from a dietician. Because of this
design it was not possible to determine the effect of nutritional education by a dietician. It
is only mentioned in Chapter 5 that adolescents found the session with the dietician
useful. In the intervention reported in Chapter 7, only one group was visited by a dietician
with the aim to test whether there were significant differences between the two text
messaging groups when adding a dietician session to one of the groups. This was not the
case and therefore it is suggested that education by a dietician in a 45-minutes session at
school is not necessary. However, Chapter 7 does not say anything about the adolescents’
perception of the session, which could be positive as reported in Chapter 5.
Another contribution in itself is the exploration of the interactivity potential
of text messaging in an intervention. Only few intervention studies on adolescent healthy
eating have tried to benefit from the popularity of text messaging among adolescents. The
finding that sending just half of the text messages – 18 daily text messages within 35 days
– can improve fruit and vegetable intake shows that text messaging has potential as a
health promotion tool. Interventions tend to be costly, but text messaging is an affordable
205
tool, available to most social groups and easy to access (Fjeldsoe, Marshall, & Miller,
2009). Hence, by exploring the potential of text messaging as a two-way communication
tool in a feedback intervention, where participants are engaged in a learning-by-doing
process to facilitate changes in behaviour, this study gives a valuable contribution to the
health promotion literature.
8.2. Limitations
This thesis has its limitations, and since most of them are inherent to the
chosen research methods and are described in more detail in each research paper, this
section will present more general limitations. First of all, the choice of focusing
specifically on fruit and vegetable intake as an important example of healthy eating
entails precautions when interpreting the results. Section 1.3 provided arguments for
operationalizing healthy eating as fruit and vegetable intake, but a study focusing on other
examples of healthy eating, such as intake of fish or wholegrain, might have generated
different results, and also comparing different healthy eating behaviours would be highly
relevant (e.g. Bruening et al. (2012)). During the work with this thesis, it has become
clear that exploring environmental and personal factors behind unhealthy eating might be
just as important as studying healthy eating, and there is a lack of knowledge about
adolescents’ unhealthy eating. Future studies should therefore look into a broader range
of behaviours, including unhealthy eating.
Secondly, studying adolescent healthy eating and aiming for a more holistic
view is challenging – especially within the confines of a doctoral study where the
empirical foundation is bound to be rather limited and interventions of short duration.
206
Since an important aim of this thesis was to deepen the understanding of the
development of and influences on an important type of healthy eating behaviour among
adolescents, it is a limitation that the qualitative study reported in the first research paper
(Chapter 5) is a cross-sectional study conducted at a single point in time. Ideally, in order
to access the consumer socialisation process in detail, the transformation of a sample of
children into adolescence should have been studied over time. Future studies might
consider, for example, longitudinal ethnographic studies. Also, the interviewed
adolescents (and in some cases also one parent) had taken part in a healthy eating
intervention three months earlier. It is a limitation of this study that the possible influence
on the findings of the participation in the intervention cannot be determined. Hence, it
cannot be rejected that a study with a similar population that had not taken part in such an
intervention could provide other results.
Thirdly, one of the main findings in this thesis is the perceived importance
of parents in adolescence in their role as main influencer on healthy eating (Chapters 5
and 6). The recommendation is to study adolescents’ healthy eating in a family context.
Also, a study on childhood obesity treatments comparing intervention effects targeting
parents alone and parents and obese children (age 6-11) together, finds that targeting
parents alone yields better results (Golan, Kaufman, & Shahar, 2006). In this connection
it may seem illogical that the intervention described in Chapter 7 does not include
parents. The decision to target adolescents only in this case was based on the findings in
the small-scale intervention study prior to the qualitative study where parents were
included, but no significant results were found. Hence, it was concluded that for this
particular intervention, it would be superfluous to also include parents actively. However,
in order to assess the apparent discrepancy between choice of design elements and
findings in this thesis, this should be explored more thoroughly in future studies.
207
Finally, the results of the intervention study reported in Chapter 7 should be
interpreted with caution: The significant effects of the level of activity on intervention
outcomes could be due to intervention design features favouring only those motivated
enough to take part. A different design paying more attention to the motivation of the
participants might have produced different results. Text message-based healthy eating
interventions among adolescents are rare, and one reason for this – partly implied by the
results reported in this thesis – could be lack of convincing results. Hence, future studies
should focus on comparing design features in search for the ‘optimal’ intervention design
and more convincing results.
8.3. Implications
8.3.1. Research implications
This PhD thesis establishes that when studying adolescent healthy eating,
both personal factors (such as self-efficacy and outcome expectations) and environmental
factors (such as social influences) should be taken into account. In Chapters 5 and 6 the
importance of parents’ healthy eating behaviour for adolescents’ healthy eating behaviour
is illustrated. This finding emphasises the importance of considering the influence – and
possibly involvement – of parents when targeting adolescents in interventions. The results
demonstrate that it is important to study family interaction since adolescents are subject to
their parents’ attitudes and behaviours regarding healthy eating. The family context is part
of the socio-structural factors directly and indirectly influencing adolescents and a setting
which to different degrees leaves room for adolescents’ direct and indirect influence. If
the adolescents’ attempts of influence are successful (for instance avoiding an unfamiliar
vegetable by suggesting a more familiar one), this could help build adolescents’ self-
208
efficacy and motivation for healthy eating – as well as motivation for taking part in
family interaction with respect to e.g. shopping and cooking. The role of motivation and
self-regulation is not new in research on healthy eating, but the direction and degree of
influence from various sources and how to build it have not been fully explored in terms
of adolescents’ healthy eating. Future research is suggested to do so.
This research also demonstrates that feedback interventions can initiate
positive behaviour changes. Chapter 7 showed how text messaging can be used as a
platform for feedback. Applying new media technology definitely has its place in health
promotion. It can transcend the boundaries between social spheres – such as home and
school – in interventions, it is not limited to one physical place or one time, and it is
relatively cheap. The reported study was rather large in scale and also close to “real-life”
in terms of voluntary participation and opt-out possibilities. The results show that when
participation levels are high enough, positive results in terms of increased intake of fruit
and vegetables can be obtained.
However, the drops in self-efficacy and outcome expectations found in
Chapter 7 for those with lower participation levels are worrying. In Chapter 7 possible
boomerang effects in terms of learned helplessness and denial are suggested as
explanations. If an intervention causes decline in self-efficacy and/or outcome
expectations, this possibly makes it more difficult for the participant to engage in other
future behaviour changes in relation to healthy eating. So it is paramount to know how to
motivate participants to participate – and participate sufficiently – in order to improve
behaviour and avoid deteriorating antecedents of behaviour. As Chapter 7 shows, just
getting participants to register can be difficult. Especially the high rate of boys (23.8 %)
not participating in the first week of the intervention is a challenge. Future studies should
look more into the barriers for registration and participation in health behaviour
209
interventions using text messaging or other interactive media. Future studies should also
explore participants’ levels of motivation at different times before and during the
intervention, the barriers for motivation and how to increase motivation. From the
experiences in this thesis, it is crucial to know more about participants’ motivation in
relation to changes in health behaviours. This research adds to the limited knowledge of
including text messaging – and new media technology in general – in health promotion
efforts. More research is needed to increase the knowledge of the importance of design
features. The literature review and the results indicate that intervention design features,
such as length of intervention and participation method (adolescent alone, parent alone or
adolescent and parent together), could also matter for the optimal intervention for
adolescents’ healthy eating. Also, it seems relevant to conduct follow up studies and
target other healthy eating behaviours as well as other health behaviours.
Finally, future research should improve the reliability of self-reported onthe-go data such as text messaging. The current knowledge is still not detailed enough to
be able to explain how and why participants report as they do, and it is important to
develop solid reliability checks.
8.3.2. Theoretical implications
This thesis demonstrates that a SCT approach to consumer socialisation
research can add to the development of this field. The chosen framework brings the two
main elements of consumer socialisation research – cognitive development theory and
social learning theory – closer together. However, it is important to develop this
integration further in future research. Especially the role of social influences and the
entanglement of social spheres should be studied in more detail in order to add to the
understanding of consumer socialisation. Although major behaviour changes are rare, the
210
SCT is widely regarded a suitable theoretical foundation for behaviour change
interventions (e.g. Michie et al., 2008). Therefore, the design of the intervention
investigated in this thesis was built on SCT, with an aim of influencing two antecedents
of behaviour: self-efficacy and outcome expectations. Despite a thorough intervention
design and careful implementation efforts, only a small – but significant – increase in
fruit and vegetable intake was found, as reported in Chapter 7. Also, despite addressing
participants’ outcome expectations and self-efficacy (through information and feedback
in the intervention), the identified behaviour change among the most active participants
was, contrary to expectations, not mediated through self-efficacy and outcome
expectations. This suggests that it is extremely difficult to change behaviour, but it also
questions whether the constructs suggested by the SCT are necessarily the most important
focus points of a healthy eating intervention.
The Health Action Process Approach (HAPA) (Schwarzer, 2008a, 2008b),
and other stage models of health behaviour (Sutton, 2005), are motivated exactly by the
somewhat disappointing results of SCT-based interventions when it comes to behaviour
change. HAPA distinguishes between pre-intentional motivation processes that form
behavioural
intentions
and
post-intentional
volition
processes
that
form
the
transformation of intentions into actual health behaviour. The volition phase is further
subdivided into three phases: planning, initiation and maintenance. Based on Bandura’s
SCT, self-efficacy and outcome expectations are assumed to influence intentions directly,
in the motivation stage. Since the found behaviour change is not mediated through these
two variables, it can be inferred that the behaviour change is (probably) not caused by a
change/improvement in motivation. Hence, it is likely that participants for some reasons
become better at acting on their motivation due to the intervention; a change in the
volition stage. The intervention may have improved participants’ planning, initiation
211
and/or maintenance with regard to eating fruit and vegetables, where improved planning
and/or initiation seems most likely. Since specific measures regarding these processes
were not taken, it is not possible to determine with any certainty which volition processes
were influenced. It is possible that goal-setting and reporting via text messages in the
intervention have stimulated the participants to concretise implementation intentions
(Gollwitzer (1993) regarding eating more fruit and vegetables (i.e., improved planning),
but the daily messages may also have functioned as prompts (i.e., initiation) and/or
reminders (i.e., maintenance) for those most active in the intervention.
The drop in self-efficacy and outcome expectations for the less active
participants can be due to their failure to live up to the challenge of the intervention.
Negative mastery experiences can make participants less sure of their self-efficacy and
make them lose faith in their capabilities (Bandura, 1997). It cannot be determined from
this research whether they will also be more prone to develop personal inefficacy
(Bandura, 1982) with regard to other behaviour changes. However, the rather optimistic
view on human’s self-motivation abilities could be questioned by these declining levels,
and even though it is on a weak empirical basis, it is a theoretical implication of this study
that there is a need to explain in detail, within the framework of SCT, how to overcome
failure and develop more resilient self-efficacy and more positive outcome expectations
as important antecedents for behaviour.
In sum, findings of this thesis question the usefulness of SCT in its original
form as a framework for feedback interventions. However, a thorough comparison of
different design features and outcomes (as suggested in section 8.2.) are needed before a
final assessment of SCT as a framework can be made.
212
8.3.3. Policy implications
There are a number of practical implications of this PhD thesis for health
promoters, such as health policy makers and other stakeholders, trying to counter the
rising levels of overweight among adolescents. In line with WHO’s recommendations
regarding securing healthy adolescents, this PhD thesis has demonstrated that it is
important to target both adolescents’ motivation and the environment surrounding them.
WHO recommends more support for parents and schools, and the findings in Chapters 5
and 6 confirm this. Still, personal skills and motivation are also important as
demonstrated in Chapter 7, where the intervention succeeded in making some of the
participants change. However, more research on the motivational aspects of engaging in
an intervention is needed and especially in order to avoid boomerang effects as the ones
described in Chapter 7. Parents’ and adolescents’ barriers and facilitators in relation to
healthy eating presented in Chapter 5 underlined the importance of a more holistic view
on adolescents’ eating behaviours, rather than focusing merely on the individual or the
environmental aspects in health policies. It is outside the scope of the thesis to suggest
specific policy changes, but it is expected that for instance taxes on healthy and unhealthy
foods play a role in what the parents need to be gatekeepers for. Furthermore, the study
found health to be important for especially parents and their awareness of being role
models for their children should therefore be reinforced, which could be done in
information campaigns. The results in Chapter 7 show that it is possible to achieve
positive results from a rather low-budget intervention by using text messaging. The
results also show that there were no significant effects of the 45-minute nutritional
education session over and above participating in the text messaging intervention. One
could claim that a 45-minute education session is a rather limited intervention, but it is
not unlike how teaching nutrition in school by an outside nutritionist typically occurs
213
(although it would probably often be given in connection with a wider class theme on
nutrition and not detached as here). Another reason for no significant effects of the 45minute education session could be that the adolescents in this study already had sufficient
knowledge about the health benefits of eating the recommended amount of fruit and
vegetables. If the participants have sufficient knowledge about healthy eating, adding an
intervention using text messaging would be more effective than adding even more
education, for example by a dietician. If pre- and post-measurements are conducted
online, it is possible – by the help of for instance schools and teachers – to conduct really
large-scale interventions at a considerably lower price than interventions demanding
many man-hours at schools or in families by dieticians. Therefore, health promoters
should consider text message-based or similar interventions for a wider population and
allocate the expertise of dieticians to those in need of help.
To sum up, this PhD thesis has contributed with knowledge about
adolescents’ healthy eating. It has deepened the understanding of the development of
healthy eating behaviours, the factors influencing healthy eating and how healthy eating
behaviours can be improved. Further research will hopefully continue to contribute with
knowledge about healthy eating for the health and benefit of future generations.
214
References
Bandura, Albert. (1982). Self-efficacy mechanism in human agency. American
Psychologist, 37(2), 122-147.
Bandura, Albert. (1986). Social foundations of thought and action: A social cognitive
perspective. Englewood Cliffs, NJ: Prentice-Hall.
Bandura, Albert. (1997). Self-efficacy: The exercise of control. New York: W. H.
Freeman.
Bech-Larsen, Tino, & Grønhøj, Alice. (2013). Promoting healthy eating to children: A
text message (SMS) feedback approach. International Journal of Consumer
Studies, 37(3), 250-256. doi: 10.1111/j.1470-6431.2012.01133.x
Bruening, M. , Eisenberg, M. , MacLehose, R. , Nanney, M. , Story, M. , & NeumarkSztainer, D. (2012). Relationship between adolescents’ and their friends’ eating
behaviors: Breakfast, fruit, vegetable, whole-grain and dairy intake. Journal of
Academy of Nutrition and Dietetics, 112, 1608-1613. doi:
10.1016/j.jand.2012.07.008
Contento, Isobel R., Randell, Jill S., & Basch, Charles E. (2002). Review and analysis of
evaluation measures used in nutrition education intervention research. Journal of
Nutrition Education and Behavior, 34(1), 2-25. doi: 10.1016/S14994046(06)60220-0
Cruwys, Tegan, Bevelander, Kirsten E, & Hermans, Roel C.J. (2015). Social modeling of
eating: A review of when and why social influence affects food intake and choice.
Appetite, 86, 3. doi: 10.1016/j.appet.2014.08.035
Fitzgerald, A., Heary, C., Kelly, C., Nixon, E., & Shevlin, M. (2013). Self-efficacy for
healthy eating and peer support for unhealthy eating are associated with
adolescents’ food intake patterns. Appetite, 63, 48-58. doi:
10.1016/j.appet.2012.12.011
Fjeldsoe, Brianna S., Marshall, Alison L., & Miller, Yvette D. (2009). Behavior change
interventions delivered by mobile telephone short-message service. American
Journal of Preventive Medicine, 36, 165-173.
Gitelson, I. , & McDermott, D. (2006). Parents and their young adult children: Transitions
to adulthood. Child Welfare, 86(5), 853-866.
Golan, Moria, Kaufman, Vered, & Shahar, Danit R. (2006). Childhood obesity treatment:
Targeting parents exclusively versus parents and children. British Journal of
Nutrition, 95(5), 1008-1015. doi: 10.1079/bjn20061757
215
Gollwitzer, Peter. (1993). Goal achievement: The role of intentions. European Review of
Social Psychology, 4, 141-185. doi: 10.1080/14792779343000059
Higgs, Suzanne. (2015). Social norms and their influence on eating behaviours. Appetite,
86(0), 38-44. doi: 10.1016/j.appet.2014.10.021
John, Deborah Roedder. (1999). Consumer socialization of children: A retrospective look
at twenty-five years of research. Journal of Consumer Research, 26(3), 183-213.
Kuczynski, L., & Parkin, M. C. (2006). Agency and bidirectionality in socialization. In J.
E. Grusec & P. D. Hastings (Eds.), Handbook of socialization: Theory and
research (pp. 259-283). New York: The Guilford Press.
Michie, Susan, Johnston, Marie, Francis, Jill, Hardeman, Wendy, & Eccles, Martin.
(2008). From theory to intervention: Mapping theoretically derived behavioural
determinants to behaviour change techniques. Applied Psychology, 57(4), 660680. doi: 10.1111/j.1464-0597.2008.00341.x
Moschis, G., & Churchill, Gilbert A. (1978). Consumer socialization: A theoretical and
empirical analysis. Journal of Marketing Research, 15, 599-609.
Nørgaard, Maria Kümpel, Brunsø, Karen, Christensen, Pia Haudrup, & Mikkelsen,
Miguel Romero. (2007). Children’s influence on and participation in the family
decision process during food buying. Young Consumers, 8(3), 197-216.
Oenema, Anke, & Brug, Johannes. (2003). Feedback strategies to raise awareness of
personal dietary intake: Results of a randomized controlled trial. Preventive
Medicine, 36(4), 429-439. doi: 10.1016/S0091-7435(02)00043-9
Rasmussen, Mette, Krolner, Rikke, Klepp, Knut-Inge, Lytle, L, Brug, Johannes, Bere, E,
& Due, Pernille. (2006). Determinants of fruit and vegetable consumption among
children and adolescents: A review of the literature. Part I: quantitative studies. Int
J Behav Nutr Phys Act, 3, 22.
Resnicow, K., Davis-Hearn, M., Smith, M., Baranowski, T., Lin, L. S., Baranowski, J., . .
. Wang, D. T. (1997). Social-cognitive predictors of fruit and vegetable intake in
children. Health Psychology, 16(3), 272-276.
Salvy, Sarah-Jeanne, Elmo, Alison, Nitecki, Lauren A, Kluczynski, Melissa A, &
Roemmich, James N. (2011). Influence of parents and friends on children's and
adolescents' food intake and food selection. The American Journal of Clinical
Nutrition, 93(1), 87-92. doi: 10.3945/ajcn.110.002097
Schwarzer, R. (2008a). Modeling health behavior change: How to predict and modify the
adoption and maintenance of health behaviors. Applied Psychology: An
International Review, 57(1), 1-29. doi: 10.1111/j.1464-0597.2007.00325.x
216
Schwarzer, R. (2008b). Models of health behaviour change: Intention as mediator or stage
as moderator? Psychology & Health, 23(3), 259-263. doi:
10.1080/08870440801889476
Sutton, Stephen. (2005). Stage theories of health behaviour. In M. Conner & P. Norman
(Eds.), Predicting health behaviour: Research and practice with social cognition
models (2nd ed., pp. 223-275). Berkshire: Open University Press.
217