1. science
2. medicine
3. psychology and psychiatry

Australian Psychologist (Accepted 19/2/14).
1
Child Adjustment and Parent Efficacy Scale (CAPES):
Development and Initial Validation of a Parent Report Measure
Alina Morawska*
Matthew R Sanders,
Divna Haslam
Ania Filus
Renee Fletcher
Parenting and Family Support Centre, School of Psychology, The University of Queensland Brisbane, Australia
*Corresponding author
Parenting and Family Support Centre
School of Psychology
University of Queensland
Brisbane 4072
Ph: +61 7 3365 7304
Fax: +61 7 3365 6724
Email: [email protected]
Abstract
Background: This study examined the psychometric characteristics of the Child Adjustment and
Parental Efficacy Scale (CAPES). The CAPES was designed as a brief outcome measure in the
evaluation of both public health and individual or group parenting interventions. The scale consists of
a 30-item Intensity scale with two subscales measuring children’s behaviour problems and emotional
maladjustment and a 20-item Self-efficacy scale which measures parent’s self-efficacy in managing
specific child problem behaviours. Method: A sample of 347 parents of 2-12 year old children
participated in the study. Results: Psychometric evaluation of the CAPES revealed that both the
Intensity and Self-efficacy scales had good internal consistency, as well as satisfactory convergent and
discriminant validity. Conclusions: Potential uses of the measure and implications for future
validation studies are discussed.
What is already known about this topic:

Child behavioural and emotional problems are common and there are a range of
effective intervention approaches.

Population level tools for assessment of child behavioural and emotional problems
are needed.

Existing tools have a number of limitations and weaknesses in application across
clinical and population levels.
What this topic adds:

This study reports on the development and piloting of the Child Adjustment and
Parent Efficacy Scale (CAPES).

The CAPES shows good psychometric properties, and therefore has the potential to
be used as a measure of child behavioural and emotional problems and parenting efficacy
across a range of contexts.

Further research is needed to establish population norms, and generalizability across
groups and cultures.
1
Australian Psychologist (Accepted 19/2/14).
2
The key to healthy optimal child development begins with positive, nurturing, and responsive
parent-child relationships (Collins, Maccoby, Steinberg, Hetherington, & Bornstein, 2000; Vimpani,
Patton, & Hayes, 2002). Traditionally, parenting programs that foster the development of such
relationships by increasing the knowledge, skills and confidence of parents have been conducted
within clinical settings targeting at-risk families or those identified with more severe child behavioural
difficulties (e.g., Sanders, Markie-Dadds, Tully, & Bor, 2000; Webster-Stratton & Reid, 2003).
However, the need for wide reaching, evidence-based parenting support is gaining increasing highlevel international recognition (e.g., Biglan, Flay, Embry, & Sandler, 2012; Sanders, 2012). To inform
public health policy and parenting interventions child adjustment measures that fulfil specific
requirements need to be developed. Such measures should have good psychometric properties, be
reliable, change sensitive, readily deployable, and be able to facilitate the tracking of intervention
outcomes at both individual and population levels.
While a number of well validated measures of child behaviour and adjustment exist, these
suffer from a number of limitations when applied in the context of a public health approach to
parenting intervention. These limitations include: (a) a focus on clinical diagnosis (e.g., Conners,
2009) and limited examination of frequent, but problematic behaviours seen among children with subclinical problems (Goodman, 1997); (b) in addition, most measures focus on older children with
limited screening tools available for children under age 5 (Bagner, Rodríguez, Blake, Linares, &
Carter, 2012; Briggs-Gowan et al., 2013). Furthermore, most measures: (c) take a deficit approach and
do not examine behavioural competencies (Tsang, Wong, & Lo, 2012); (d) a focus on only one type
of behaviour (e.g., Kovacs, 2010) or only on internalizing (e.g., Spence, 1998) or externalizing (e.g.,
Eyberg & Pincus, 1999) behaviour. The other problematic issues with the existing measures include:
(e) low to moderate estimates of internal consistency (e.g., Strenghts and Diffiuclites Questionnaires
.51-.76; Smedje, Broman, Hetta, & von Knorring, 1999); (f) measures such as the CBCL (Achenbach,
2000) that are lengthy and time consuming for parents to complete (Goodman & Scott, 1999), and; (g)
licensing fees, which can place limits on dissemination in large population studies (e.g., Eyberg &
Pincus, 1999; Goodman, 1997). In particular, the Strengths and Difficulties Questionnaire which is a
widely used screening tool, suffers from a number of limitations for population level use, such as
relatively low internal consistency (Goodman, 2001) particularly for some subscales (e.g., peer
problems subscale), limited data for children younger than seven years (Mieloo et al., 2012),
questions about the underlying structural validity (Palmieri & Smith, 2007), limited data on sensitivity
to change (Tsang et al., 2012), as well as the requirement of a licence fee for online administration.
Furthermore, in the context of parenting and parenting intervention, parental self-efficacy is
increasingly being recognised as important to understanding the ways in which parent-child
relationships and child behavioural and emotional problems develop and maintain over time (Jones &
Prinz, 2005) and none of the existing measures include parental self-efficacy in relation to the specific
behavioural and emotional problems measured.
The term “self-efficacy”, often used interchangeably with “confidence”, is defined as “the
conviction that one can successfully execute the behaviour required to produce the outcomes”
(Bandura, 1977). Although “confidence” refers to the strength of a particular belief, the construct of
self-efficacy, as defined by Bandura, specifically pertains to an individual’s belief that they can
perform a given activity successfully, as well as to the strength of that belief (Bandura, 1997). Thus,
self-efficacy beliefs are attached to specific domains of functioning (e.g. parenting) (Bandura, 2000).
Parental self-efficacy has been shown to impact on children’s behaviour both directly and indirectly
via parenting practices, and on parental adjustment (Jones & Prinz, 2005). Furthermore, it has been
identified as an important modifiable factor for parenting intervention (Morawska & Sanders, 2007)
alongside parental competence (Jones & Prinz, 2005). While there are a number of parental efficacy
measures, these are generally measures of global self-efficacy (e.g., Johnston & Mash, 1989), despite
the fact that task specific efficacy is a better predictor of child outcomes (Jones & Prinz, 2005).
Similarly, while task-specific measures do exist (e.g., Sanders & Woolley, 2005) these are not
integrated with existing measures of child behavioural problems, thus increasing the assessment
burden on families when both variables are assessed. Furthermore, there may be differences between
a parent’s perceptions of their child’s behaviour and their own self-efficacy in dealing with the
behaviour which can be important to differentiate in an intervention approach. For example, the
2
Australian Psychologist (Accepted 19/2/14).
3
parent who experiences relatively low levels of behaviour problems, but nevertheless has low selfefficacy about managing these might need quite a different treatment approach to one who
experiences high levels of behaviour problems and has high self-efficacy.
The current study provides initial validation of a newly developed measure of child
behavioural and emotional adjustment that would overcome some of the limitations outlined above,
enabling population level assessment of child adjustment, and facilitating outcome evaluation and
meta-analysis of population-level interventions. In developing the assessment tool we sought to create
a measure that would meet a number of different criteria. Specifically, the measure needed to be easy
to administer, score, and interpret, use consistent methods of scaling, and have sound psychometric
properties in terms of reliability, face and construct validity. In addition, we sought to develop a
measure that would have the following characteristics: (a) minimal assessment burden, (b) be
sensitive to change in both clinical and non-clinical populations, (c) be comprehensive enough to
cover primary targets of parenting interventions focusing on improving children’s adjustment
(reducing externalising and internalising problems and increasing prosocial and adaptive behaviours),
(d) be applicable for interventions of different intensities and, (e) have the potential to be used as a
population level indicator of the prevalence of behavioural and emotional problems of children. While
we designed the measure to address a number of limitations in the literature, this study provides the
results of an initial validation of this measure, and serves as the beginning of a broader program of
research to evaluate the measure. We aimed to test the psychometric properties of the Child
Adjustment and Parent Efficacy Scale (CAPES). Specifically we sought to: 1) apply principles of
measure development to create a brief parent report, user friendly, public domain measure of
children’s adjustment and parental self-efficacy in managing children’s behaviour; 2) determine the
construct validity of the CAPES, and; 3) determine the internal consistency of the scale.
Method
Participants
The sample consisted of 347 parents of children 2-12 years old who were recruited Australiawide from schools and day care centres, online forums, and parenting newsletters. The only eligibility
criteria were that parents have a child between ages 2-12 years. Parents’ ages ranged from 24 to 58
years (M=39.49, SD=5.98). The majority of the sample self-identified as Caucasian/Australian
(n=250, 72.1%) with the remaining identifying as Aboriginal and Torres Strait Islander (n=7, 2.0%),
Asian (n=5, 1.4%), or other (n=8, 2.3%)1. More mothers (n=295, 85.0%) than fathers (n=14, 4.0%)
responded to the questionnaire. Children’s ages ranged from 2 to 12 years (M=7.34, SD=2.80) and
there were more girls (n=180, 51.9%) than boys (n=129, 37.2%). A range of parental education
attainment was represented with 166 (47.8%) having a university degree, 76 (21.9%) completing part
or all of high school, and 67 (19.3%) completing trade or technical college. Most parents were
married (n=229, 66.0%) and employed (n=239, 68.9%). The majority of parents (n=240, 69.2%)
reported having no difficulties meeting essential household expenses, whereas 67 (19.3%) declared
having problems meeting essential expenses over the last 12 months. Furthermore, 104 (30.0%)
reported that they earn enough to comfortably purchase most of the things they really want, 146
parents (42.1%) declared that their earnings allow them to purchase only some things that they want,
while 59 parents (17.0%) reporting they don’t have enough money to purchase much of anything they
really want.
Procedure
Ethical clearance for the study was obtained in accordance with the ethical review processes
of the University of Queensland. The following steps were taken in designing the measure: (1)
definition of constructs; (2) review of existing measures; (3) generation of initial item pool; (4) input
and feedback from key experts; (5) input and feedback from parents, and; (6) initial piloting to assess
psychometric properties. In determining the construct domains for assessment, we focused on
1
Numbers do not add to 100% due to missing data.
3
Australian Psychologist (Accepted 19/2/14).
4
behavioural difficulties and anxiety, as these are the most prevalent types of difficulties experienced
by children (Morawska & Sanders, 2011), and on child strengths and competencies to be consistent
with interventions designed to improve child skills and self-regulatory abilities (e.g., Sanders, 2012).
Furthermore, we wanted to integrate ratings of parenting efficacy which were domain specific, given
the increasing focus on task-specific self-efficacy in the parenting literature (Jones & Prinz, 2005).
We reviewed existing validated measures, including examining our own data from a range of
intervention and population studies (e.g., Morawska & Sanders, 2006; Sanders et al., 2008) to identify
common difficulties experienced by parents. The initial item pool was generated on the basis of this
review in the context of our focus on child competencies. The initial scale was disseminated to a
number of international experts in the parenting literature for feedback and to ensure wording and
content were culturally relevant. Four in-depth semi-structured interviews were conducted with
parents to gain feedback about ease of understanding, the completion process and face validity.
Parents were asked to complete the questionnaire like they usually would and then to use a highlighter
to mark anything on the questionnaire that was unclear or ambiguous. They were then asked a series
of questions designed to elicit feedback (e.g., Is there anything that would make the survey easier to
complete? Is there anything missing from the questionnaire that is important to you?). Feedback from
the interviews indicated the measure was easy to understand and had high face validity however some
parents recommended rewording items slightly. Several items were modified in response to expert
and parent feedback, the order of items was changed and some items were dropped. This resulted in
the final 30-item Child Adjustment and Parent Efficacy Scale (CAPES; Morawska & Sanders, 2010,
see Appendix A).
To ensure the revised measure could be understood by a wide range of parents it was assessed
for readability using the Flesch reading ease and the Flesch-Kincaid grade tests. These tests assess
comprehension difficulty and provide an estimate of education grade level (grade 1- 12) required for
understanding. Scores of 65.1 (out of a possible 100 where higher scores indicate greater ease) and 9.5
(possible range 1-12) were obtained on the Flesch reading ease test and the Flesch-Kincaid grade tests
respectively indicating the measure could be easily understood by a student aged 13-15 years or
someone with a grade 9 level education.
Following the measure development and consultation process initial piloting was conducted
to assess the psychometric properties of the measure. An online survey was created and a large parent
sample was recruited to complete the questionnaire. Recruitment was via school and childcare
newsletters, online posts at parenting websites and via media releases. The number of parents who
may have seen information about the survey but chose not to participate is unknown. Parents were
directed to a website where they read a brief information sheet and provided informed consent prior to
completing the questionnaire anonymously.
Measures
The Family Background Questionnaire (Sanders & Morawska, 2010) was used to assess
family demographic characteristics, including child and parent age and gender, family composition,
parent marital status, ethnicity and education and income.
The Child Adjustment and Parent Efficacy Scale (CAPES; Morawska & Sanders, 2010) is a
measure of child behavioural and emotional adjustment and parental efficacy. It consists of 30 items
rated on a 4-point scale, ranging from not true of my child at all (0) to true of my child very much, or
most of the time (3), where 20 items are two part questions that assess both child behaviour and parent
efficacy. Twenty-six items assess behaviour concerns (e.g., My child rudely answers back to me) and
behavioural competencies (Behaviour Scale; e.g., My child follows rules and limits), and four items
assess emotional adjustment (Emotional Maladjustment Scale; e.g., My child worries). Some items
are reverse scored. Items are summed to yield a total intensity score (Capes Intensity Scale: range of
0-90), which is made up of a behaviour score (range of 0-78) and an emotional maladjustment score
(0-12) where high scores indicate higher levels of problems. The Self-efficacy Scale consists of 20
items and measures parents’ level of self-efficacy in managing child emotional and behavioural
problems. Items are rated on a 10-point scale, ranging from certain I can't do it (1) to certain I can do
4
Australian Psychologist (Accepted 19/2/14).
5
it (10). A total efficacy score with a possible range of 20-200 is calculated by summing all efficacy
items, with higher scores indicating a greater level of self-efficacy.
Analytical Procedure
Construct validity refers to the extent to which the scale measures what it is supposed to
measure (Nunnally & Bernstein, 1994). In order to establish construct validity the factor structures of
the CAPES Intensity and Self-efficacy were examined through confirmatory factor analysis (CFA)
using AMOS v.20. The maximum likelihood estimation method was applied to the analysis of the
covariance matrices2. The absolute goodness-of-fit of the models was evaluated using the χ2 test and
three additional criteria: the comparative fit index (CFI), the root mean square error of approximation
(RMSEA) with 90% confidence interval, and the standardized root mean square residual (SRMR).
The CFI index is a revised version of the Bentler and Bonett (1980) normed fit index that adjusts for
degrees of freedom; values above .90 are considered adequate and above .95 as very good (Hu &
Bentler, 1999). The RMSEA index indicates the error of approximation; values of less than .05 are
considered good, though values as high as .08 are also considered reasonable (Browne & Cudeck,
1989). The SRMR index is an absolute measure of fit and represents the difference between the
observed correlation and the predicted correlation; values close to .08 or less represent a good fit (Hu
& Bentler, 1999).
Models were respecified based on Modification Indices (MIs), inspection of standardized
residuals and theoretical considerations (Kline, 2011). The MIs values indicate the expected decrease
in χ2 given the relaxation of imposed constraints; values < 5.00 indicate little appreciable improvement
fit. A particularly high standardized residual for the covariance between two variables indicates that
the relationship between those variables is not well accounted for by the model. In addition, to assess
the extent to which a newly specified model exhibits an improvement over its predecessor, the χ2
difference test (Δχ2) was used. A significant Δχ2 indicates a substantial improvement in model fit.
Convergent validity refers to the degree to which a set of measurement items appear to be indicators
of one single underlying latent variable. Three approaches were applied to assess convergent validity:
(1) we examined whether factor loadings for each indicator were statistically significant (Gerbing &
Anderson, 1988); (2) checked that the estimate of the average variance extracted (AVE) that is shared
between the construct and its measures is above .50 (Fornell & Larcker, 1981)3, and; (3) tested that
estimates of composite reliability (CR) were above .704.
Three approaches were also used to assess the discriminant validity of the measure.
Discriminant validity refers to the extent to which items that measure one construct are different from
the items that measure another construct. First, we analysed the correlations between the latent
constructs, which should not be close or equal to 1.00. As an extension to this method we used a χ2
difference test (Bollen, 1989). In this test a model is analysed, in which the correlation between the
factors is fixed at 1.00. The constrained model’s χ2 is compared to the original model’s χ2 where the
correlation between the constructs is estimated freely. Discriminant validity is shown when the
unconstrained model has a significantly lower chi-square value, indicating that the constructs are not
perfectly correlated. The third method included the comparison of AVE to the squared interconstruct
2
The variance-covariance matrices available on request from the corresponding author.
3
The AVE estimate represents the average amount of variation that a latent construct is able to explain in the
observed variables that theoretically relate to the construct. It is calculated by averaging the sum of squared
factor loadings for each latent construct. The squared factor loading represents the amount of variation in each
observed variable that the latent construct accounts for. When this variance is averaged across all observed
variables that relate theoretically to the latent construct, we generate the AVE.
4
The composite reliability represents the overall reliability of a collection of heterogeneous yet similar items. It
reflects the degree to which the scale score reflects one particular factor. It is calculated in the following way:
CR = (sum of standardized factor loadings)2/ (sum of factor loadings)2 + (sum of indicator measurement error).
5
Australian Psychologist (Accepted 19/2/14).
6
correlation estimates (SIC); the AVE for each construct should be larger than the SIC among these
constructs (Fornell & Larcker, 1981).
Internal consistency of CAPES Intensity and CAPES Self-efficacy was examined using
Cronbach’s alphas computed in SPSS v.20. Values above .70 were considered good indicators of
internal consistency (De Vaus, 2002)
Results
Preliminary Analyses
Three hundred and seventy parents responded to the survey, but 23 of these provided only
demographic data and did not complete the questionnaires; hence, they were excluded from analyses.
For the remaining 347 respondents those who did not respond to any of the Intensity (n=1, 0.2%) or
Self-efficacy (n=66, 19.0%) items were also excluded from analyses. This gave a total sample of
n=346 for Intensity and n=281 for Self-efficacy. The expectation-maximization method in SPSS v.20
was used to handle missing data (<1.5% for both Intensity and Self-efficacy).
A minimal recommended sample size in SEM studies is 200 cases (Kline, 2011). Recent
simulation studies indicate that the recommended sample sizes for confirmatory factor analysis are N
≥ 200 for theoretical models and N ≥ 300 for the population models (Myers, Ahn, & Jin, 2011) In
addition, empirical research of MacCallum, Widaman, Zhang, and Hong (1999) suggests that the
adequacy of factor analysis results depends more on data characteristics (e.g. communalities) than on
the sample size employed. When the communalities are high the sample size can be smaller. As can
be seen in Figures 1-2, the communalities for the tested scales were moderate to high in size. Under
these guidelines the available sample of 346 participants for CAPES Intensity and 281 participants for
CAPES Self-efficacy are acceptable for testing models presented in Figures 1-2.
Data were examined for departures from both univariate and multivariate normality, and for
the presence of potential outliers. For 10 out of 30 Intensity items skewness and kurtosis estimates
exceed the absolute value of 1 (average skewness and kurtosis were 0.77 and 0.36, respectively). The
normalized estimate of Mardia’s coefficient of multivariate kurtosis was 93.42 with the Critical Ratio
(C.R.) value of 19.83 indicating multivariate non-normality of the sample (Bentler, 2005; Byrne,
2010). In addition, univariate outliers were detected; as a result 277 (2.7%) extreme data points were
transformed by changing the value to the next highest/lowest (non-outlier) number. A review of
squared Mahalanobis distances (D2) showed minimal evidence of serious multivariate outliers (Byrne,
2010). In the case of Self-efficacy, for 14 out of 20 items skewness and kurtosis estimates exceed the
absolute value of 1 (average skewness and kurtosis were -1.12 and 0.66, respectively). The
normalized estimate of Mardia’s coefficient of multivariate kurtosis was 240.31 with C.R. value of
67.90 implying multivariate non-normality of the sample (Bentler, 2005; Byrne, 2010). Univariate
outliers were also detected, and thus 151 (2.7%) of extreme data points were transformed by changing
the value to the next highest/lowest (non-outlier) number. A review of D2 showed minimal evidence
of serious multivariate outliers (Byrne, 2010).
The deviation from normality of many items violated the assumptions on which normal
theory maximum likelihood estimation technique is based. Thus, the factor validity of CAPES
Intensity and Self-efficacy was assessed using a parcelling approach. A parcelling approach reduces
bias caused by non-normal distributions and increases the stability of parameter estimates in
a complex model (Coffman & MacCallum, 2005; Kishton & Widaman, 1994).
Confirmatory Factor Analysis
CFA using individual items was conducted to examine unidimensionality of the items and as
a prerequisite to the subsequent parcelling. Examination of factor structure coefficients showed that
for both CAPES Intensity and Self-efficacy none of the items exhibited problems; each item had a
high and significant factor loading on its hypothesized factor.
Construct validity evidence for Intensity and Self-efficacy was assessed using parcels-based CFA. At
least 2 items (two observed variables) are needed to build a parcel. For adequate model identification
at least three indicators are necessary within each factor (Kline, 2011), and as the Emotional
maladjustment scale consists of four items, parcelling would violate this assumption. Four observed
6
Australian Psychologist (Accepted 19/2/14).
7
variables of the Emotional maladjustment subscale would be reduced to 2 parcels (2 items per parcel),
which would cause identification problems. Therefore, for CAPES Intensity only items within the
Behaviour subscale were parcelled. Each parcel represented an average of the two individual items.
Items were combined in such a way as to maximize the normality of the resulting parcels. The
detailed information on which items were combined into which parcels is listed in the footnote5.
Validity of CAPES Intensity. The analysis of the factor structure of CAPES Intensity started
with a single model factor (Model A) to serve as a comparison to the hypothesized two factor model.
As Table 1 shows, the overall fit of the one-factor model to the data was poor, therefore it was
rejected. In the next step a model with two correlated factors named Emotional maladjustment and
Behaviour was tested (Model B). Model B showed much better fit to the data compared to Model A
according to all fit indices. In addition, the change in the χ2 value between the one and two-factor
model was significant, which also indicated that the two-factor model was adequate. Therefore we
tested the two-factor model.
As presented in Table 1, Model B achieved acceptable fit according to the SRMS and
RMSEA, but not according to χ2 and CFI. A statistically significant lack of fit as indexed by χ2
difference test is quite typical in social sciences research because of the sensitivity of the test in large
samples or with many degrees of freedom (Byrne, 2010). Nevertheless, the CFI index showed that the
model could be improved.
Inspection of standardized residuals indicated that Model B did not adequately account for the
associations between Parcel 5 (Item 12 “My child misbehaves at school or day care” & Item 28 “My
child gets on well with other children”) and Item 27 (“My child seems to feel good about
him/herself’). To assess the extent to which these items contaminate the factor validity of CAPES
Intensity, a second model (Model C) was specified, in which Parcel 5 and Item 16 were deleted. As
expected, results demonstrated a significant improvement in fit (see Table 1). However, the CFI value
below .95 indicated that a certain degree of model misfit remained. A review of MIs revealed that the
model fit could be improved by allowing several of the error terms to correlate. These modifications
were made one at a time until the fit indices met the cut-off criteria for acceptable levels. All
correlations between error terms were theoretically sound. In Model D we allowed the correlations
between error terms of Parcel 2 (Items 2 and 16) and Parcel 12 (Items 6 and 10), both containing
items referring to child’s non-compliance and disobedience. In Model E we allowed the correlation
between error terms of Parcel 3 (Items 4 and 14) and Parcel 13 (Items 23 and 30), both comprised of
items referring to child’s tantrum outburst and immaturity. Finally in Model F we allowed correlation
between error terms of Parcel 9 (Items 25 and 26) and Parcel 11 (Items 8 and 20), both comprising of
items referring to personal independence and social ability with others.
As shown in Table 1, fit values based on CFI, SRMR and RMSEA indicated that a reasonable
amount of data was explained by Model F [χ2(86)=174.88, p<.001; CFI=.952; SRMR=.052;
RMSEA=.055 (90% CI .043-.066)]. The chi-square difference between Model E and Model F was
8.99 (df=1) indicating a significant improvement (p<.001) of model fit. Thus, further fitting of the
model would clearly represent an overfit (Byrne, 2010). Therefore, Model F was selected as an
adequate description of the data. A graphic illustration of the final model is presented in Figure 1.
The AVE estimate for the Behaviour scale reached the value of .40, slightly lower than the
recommended cut-off value of .50. For Emotional maladjustment the AVE estimate reached the value
of .51, exceeding the recommended cut-off value of. 50. The CR estimates for both, Behaviour and
Emotional maladjustment were above .70 (.95 and .85, respectively). In addition, all items and parcels
5
For the CAPES Intensity, Behavior subscale: Parcel 1 included items 24 & 1, Parcel 2 included items 2 &16,
Parcel 3 included items 4 &14, Parcel 4 included items 5 &7, Parcel 5 included items 12 &28, Parcel 6 included
items 9 &18, Parcel 7 included items 21 &13, Parcel 8 included items 15 &17, Parcel 9 included items 25 &26,
Parcel 10 included items 22 &29, Parcel 11 included items 8 &20, Parcel 12 included items 6 &10, Parcel 13
included items 23 &30. For the CAPES Confidence: Parcel 1 included items 1 &2, Parcel 2 included items 3
&4, Parcel 3 included items 5 &6, Parcel 4 included items 7 &8, Parcel 5 included items 9 &10, Parcel 6
included items 11 &12, Parcel 7 included items 12 &16, Parcel 8 included items 13 &14, Parcel 9 included
items 17 &18, Parcel 10 included items 19 &20.
7
Australian Psychologist (Accepted 19/2/14).
8
had significant loadings >.40 on the factors they were specified to measure (Stevens, 1992). The
lowest factor loading on the Behaviour scale was .45 and the highest was .81. The factor loadings on
the Emotional maladjustment scale ranged from .67 to .79. These results indicate that CAPES
Intensity exhibited adequate convergent validity.
The correlation between the two factors was moderate (r=.36, p<.001). A model with fixed
correlation between the two factors was compared to one in which correlation was freely estimated.
The two models differed significantly [Δχ2(1)=386.96-174.88=212.08, p<.001], indicating that the
better model is the one in which the two constructs are viewed as distinct, yet correlated factors. The
SIC between the two factors reached the value of .13 and was much lower than the AVE estimates for
both Behaviour (.40) and Emotional maladjustment (.51). These results imply strong evidence for
discriminant validity of the two-factor model of CAPES Intensity.
Validity of CAPES Self-efficacy. The analysis of the factor structure of CAPES Self-efficacy
started with the hypothesized single factor model (Model A). As Table 2 shows, Model A achieved
acceptable fit according to the CFI and SRMS, but not according to χ2 and RMSEA.
Inspection of standardized residuals indicated that Model A adequately accounted for the
associations between the variables. A review of MIs revealed that the model fit could be improved by
allowing several of the parcels’ errors to correlate. These modifications were made one at time until
the fit indices met the cut-off criteria for acceptable levels. All correlations between error terms were
theoretically sound. In Model B we allowed the correlations between error terms of Parcel 6 (Items 11
and 12) and Parcel 10 (Items 19 and 20), both containing items referring to child fears and anxieties as
well as conduct problems. In Model C we allowed the correlation between error terms of Parcel 2
(Items 3 and 4) and Parcel 7 (Items 12 and 16), both comprised of items referring to child’s tantrum
outburst and misbehaviour at school and at home. Finally in Model D we allowed correlation between
error terms of Parcel 3 (Items 5 and 6) and Parcel 4 (Items 7 and 8), both comprising of items
referring to child’s conduct problems during mealtimes and dressing up.
As shown in Table 2 fit values based on CFI, SRMR and RMSEA indicated that a reasonable
amount of data was explained by Model D [χ2(32)=89.06, p<.001; CFI=.980; SRMR=.022;
RMSEA=.080 (90%CI .060-.100)]. The chi-square difference between Model D and Model E was
12.92 (df=1) indicating a significant improvement (p<.001) of model fit. Thus, further fitting of the
model would clearly represent an overfit (Byrne, 2010). Therefore Model D was selected as an
adequate description of the data. A graphic illustration of the final model is presented in Figure 2.
The AVE estimate for CAPES Self-efficacy reached the value of .72, indicating sufficient
amount of variance explained by the construct. The CR estimate for the scale reached the value of .88
indicating good internal consistency of the construct. Furthermore, all parcels had significant loadings
>.40 (range from .75 to .89) on the hypothesized factor (Stevens, 1992). These results indicate good
convergent validity for CAPES Self-efficacy.
As CAPES Self-efficacy is unifactorial, the Self-efficacy construct was compared against the
Behaviour and Emotional maladjustment constructs of CAPES Intensity. The two scales were tested
in one model. The correlations between Self-efficacy and Behaviour and Emotional maladjustment
constructs were moderate (r=-.33 and r=-.65, p<.001, respectively). The chi-square difference tests
indicated that discriminant validity was achieved between Self-efficacy and Behaviour
[Δχ2(1)=1263.32-878.68=384.64, p<.001] and between Self-efficacy and Emotional maladjustment
[Δχ2(1)=1046.60-878.68=167.92, p<.001]. The SIC reached the value of .11 for the relation with the
Behaviour construct, which is lower than the AVE estimates for both Behaviour (.40) and Selfefficacy (.72). Further, the SIC estimate reached the value of .42 for the relation with the Emotional
maladjustment construct, which is lower than the AVE estimates for both Emotional maladjustment
(.51) and Self-efficacy (.72). These results imply strong evidence for discriminant validity of the onefactor measurement model of CAPES Self-efficacy.
Reliability
For CAPES Intensity Cronbach’s alphas were .90 (Total scale score), .90 (Behaviour) and .74
(Emotional maladjustment). The Cronbach’s alpha for CAPES Self-efficacy was .96. Thus,
8
Australian Psychologist (Accepted 19/2/14).
9
Cronbach’s alphas were above the recommended cut-off value of .70 for good internal consistency of
the measure (De Vaus, 2002).
Discussion
The present study used an expert informant approach to construct a brief, easy to administer,
valid measure of children’s adjustment that could be used in both population-based and individual
clinical outcome evaluations of parenting interventions. We also sought to determine the
psychometric properties of the CAPES as a measure of child adjustment including the validation of
the factor structure and reliability of the scale, and a measure of parental self-efficacy in managing
children’s problem behaviours.
With respect to the measure development process we harnessed the views of experienced
international parenting researchers and clinicians to identify an item pool of 30 items that would
provide a brief, valid measure of both externalizing and internalizing difficulties that could be
subjected to psychometric evaluation. The resulting scale has the advantage of including items that
address both externalizing and internalizing problems as well as prosocial behaviours of children. The
relatively wide age range 2-12 for the scale means it could potentially be used in longitudinal studies
or in follow up evaluations of early intervention programs targeting families. The readability of the
scale indicated it could be comfortably read by parents with a grade 9 education level.
Both the CAPES Intensity and Self-efficacy scales proved to be psychometrically sound
measures. Confirmatory factor analysis supported a 27-item, two-subscale structure of the CAPES
Intensity. The findings suggested that three items require either deletion or content modification. Two
of these items, 12“My child misbehaves at school or day care” and 28 “My child gets on well with
other children” were designed to measure the child’s behaviour problems. These items may not have
fit well because they reflect behaviour out of the home setting. Item 27 “My child seems to feel good
about him/herself’ may not be a valid indicator of child’s emotional adjustment.
Furthermore, CAPES Intensity showed reasonable convergent validity as measured by
average variance extracted estimates, the composite reliability estimates and examination of factor
loadings. In other words, the CAPES Intensity items appear to be good indicators of the two
constructs, Behaviour and Emotional maladjustment, they were designed to measure. In terms of
discriminant validity, the results confirmed that the two subscales of CAPES Intensity, Behaviour and
Emotional maladjustment represent two distinct, yet correlated constructs. Evidence also suggested
that CAPES Intensity is an internally consistent measure.
As far as CAPES Self-efficacy is concerned, confirmatory factor analysis supported a 20-item,
1-factor structure of the scale. Regarding construct validity, CAPES Self-efficacy showed good
convergent and discriminant validity. The results indicated that the 20 items of CAPES Self-efficacy
appear to be indicators of the one latent construct - parental self-efficacy. To test the discriminant
validity of CAPES Self-efficacy its factor structure was compared against the Behaviour and
Emotional maladjustment constructs of CAPES Intensity. The analyses indicated that parental Selfefficacy represents a distinct construct to both the Behaviour and Emotional maladjustment constructs
of CAPES Intensity. Moreover, CAPES Self-efficacy demonstrated excellent internal consistency.
The advantage of having several developmentally important constructs related to children’s
social competence, mental health and wellbeing (e,g., conduct problems, emotional difficulties,
prosocial behaviour) and a construct related to the quality of the family environment (Viz parental
efficacy) in the one scale is that these problems can be concurrently assessed and potential targets for
both child and parental intervention identified. In evidence based parenting and family based
prevention and treatment programs parental self-efficacy is often a primary target of intervention and
changes in parental self-efficacy are frequently associated with reductions in children’s behavioural
and emotional problems (Sanders & Mazzucchelli, 2012).
This study provides the first evidence that the psychometric properties of the CAPES make it
a potentially useful measure, and serves as the basis for a more extensive evaluation of the measure
and application across a range of contexts. The findings of this initial validation study need to take
into account a number of limitations. Firstly, while the sample size was reasonable, we did not
specifically include clinical cases. Future studies should examine more heterogeneous groups and
9
Australian Psychologist (Accepted 19/2/14).
10
include multiple-group analysis to examine equivalence between clinical and non-clinical populations,
mothers and fathers, boys and girls as well as different age groups of children. As the initial
psychometric evaluation was promising further research is warranted to establish age and gender specific norms, and cut offs to determine clinical caseness of the CAPES Intensity scale when used
with children with or at risk of serious mental health problems. In addition, the current Emotional
maladjustment subscale is made up of a very small number of items, and indeed there is limited
assessment of child sadness and anxiety. Further extension of this subscale is required, particularly as
several items were removed during the initial piloting and subsequent analysis stages.
Notwithstanding these limitations further psychometric investigation of the scale in the
context of prevention and treatment studies are required to establish the change sensitivity of the
measure in trials of parenting programs. An initial study by Morawska, Tometzki and Sanders (2012)
examining this possibility showed that the CAPES scores successfully captured improvements in
children conduct problems following parents’ participation in a version of the Triple P-Positive
Parenting Program delivered as a radio podcast intervention. The validation also needs to be extended
to more diverse samples in terms of sex, age and ethnicity. Further, there is a need for evaluation of
the validity of CAPES by investigating patterns of relationships between CAPES and other measures
assessing similar and different constructs. Research is also warranted to examine the cross cultural
robustness of the scale. Finally, studies examining the convergent and discriminant validity of the
CAPES would be useful to determine the extent to which the intensity scale correlates with other
measures of the same constructs including independent behavioural observations of child behaviour,
parenting behaviour, parent-child interaction and teacher assessment of behavioural and emotional
problems.
The CAPES has good potential both as a clinical and research tool and is currently being used
and evaluated across varying contexts and cultures where is it demonstrating good psychometric
properties (e.g., Mejia, Filus, Calam, Morawska, & Sanders, 2014; Sumargi, Sofronoff, & Morawska,
2013). Its strength lies in its brevity and assessment of both child behaviour and parenting efficacy in
a single measure, which has significant benefits in assessment in a clinical setting where a
psychologist may need to assess a variety of areas to effectively formulate and to offer a tailored
intervention.
10
Australian Psychologist (Accepted 19/2/14).
11
References
Achenbach, T. M. (2000). Child Behavior Checklist 1½-5. Burlington, VT: University of Vermont
Department of Psychiatry.
Bagner, D. M., Rodríguez, G. M., Blake, C. A., Linares, D., & Carter, A. S. (2012).
Assessment of behavioral and emotional problems in infancy: A systematic review. Clinical Child
and Family Psychology Review, 15(2), 113-128. doi: 10.1007/s10567-012-0110-2
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84(2), 191-215.
Bandura, A. (1997). Self efficacy: the exercise of control. New York: W.H. Freeman and Company.
Bandura, A. (2000). Self-efficacy: the foundation of agency. In W. J. Perrig & A.Grob (Eds.), Control
of human behavior, mental processes, and consciousness. Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Bentler, P. M. (2005). EQS 6 Structural equations program manual. Encino, CA: Multivariate
Software.
Bentler, P. M., & Bonett, D. G. (1980). Significance Tests and Goodness of Fit in the Analysis of
Covariance-Structures. Psychological Bulletin, 88(3), 588-606. doi: 10.1037/00332909.88.3.588
Biglan, A., Flay, B. R., Embry, D. D., & Sandler, I. N. (2012). The Critical Role of Nurturing
Environments for Promoting Human Well-Being. American Psychologist, 67(4), 257–271. doi:
10.1037/a0026796
Bollen, K. A. (1989). Structural equations with latent variables. New York: Wiley.
Briggs-Gowan, M. J., Carter, A. S., McCarthy, K., Augustyn, M., Caronna, E., & Clark, R. (2013).
Clinical Validity of a Brief Measure of Early Childhood Social–Emotional/Behavioral
Problems. Journal of Pediatric Psychology, 38(5), 577-587. doi: 10.1093/jpepsy/jst014
Browne, M. W., & Cudeck, R. (1989). Single Sample Cross-Validation Indexes for CovarianceStructures.Multivariate
Behavioral
Research,
24(4),
445-455.
doi:
10.1207/s15327906mbr2404_4
Byrne, B. M. (2010). Structural equation modeling with AMOS: Basic concepts,applications, and
programming. Mahwah, NJ US: Lawrence Erlbaum Associates Publishers.
Coffman, D. L., & MacCallum, R. C. (2005). Using parcels to convert path analysis models into latent
variable models. Multivariate Behavioral Research, 40(2), 235-259.
Collins, W. A., Maccoby, E. E., Steinberg, L., Hetherington, E. M., & Bornstein, M.H. (2000).
Contemporary research on parenting: The case for nature and nurture. American Psychologist,
55, 218-232.
Conners, C. K. (2009). Conners Rating Scales-Revised (CRS-R). Toronto, Ontario, Canada: MultiHealth Systems.
De Vaus, D. A. (2002). Analyzing social science data. London: SAGE.
Eyberg, S. M., & Pincus, D. (1999). Eyberg Child Behavior Inventory and Sutter-Eyberg Student
Behavior Inventory - Revised: Professional manual. Odessa, FL: Psychological Assessment
Resources.
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable
variables and measurement error. Journal of Marketing Research, 39-50. doi: 10.2307/3151312
Gerbing, D. W., & Anderson, J. C. (1988). An Updated Paradigm for Scale Development
Incorporating Unidimensionality and Its Assessment. Journal of Marketing Research, 25(2),
186-192. doi: 10.2307/3172650
Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A research note.
Journal of
Child Psychology and Psychiatry, 38, 581-586.
Goodman, R. (2001). Psychometric Properties of the Strengths and Difficulties Questionnaire.
Journal of the American Academy of Child and Adolescent Psychiatry, 40(11), 1337-1345.
Goodman, R., & Scott, S. (1999). Comparing the Strengths and Difficulties Questionnaire and the
Child Behavior Checklist: Is Small Beautiful? Journal of Abnormal Child Psychology, 27(1),
17-24. doi: 10.1023/a:1022658222914
11
Australian Psychologist (Accepted 19/2/14).
12
Hu, L. T., & Bentler, P. M. (1999). Cutoff Criteria for Fit Indexes in Covariance Structure Analysis:
Conventional Criteria Versus New Alternatives. Structural Equation Modeling-a
Multidisciplinary Journal, 6(1), 1-55. doi: Doi 10.1080/10705519909540118
Johnston, C., & Mash, E. J. (1989). A measure of parenting satisfaction and efficacy.
Journal of Clinical Child Psychology, 18, 167-175.
Jones, T. L., & Prinz, R. J. (2005). Potential roles of parental self-efficacy in parent and child
adjustment: A review. Clinical Psychology Review, 25(3), 341-363 doi:
10.1016/j.cpr.2004.12.004
Kishton, J. M., & Widaman, K. F. (1994). Unidimensional Versus Domain
Representative
Parceling of Questionnaire Items - an Empirical Example. Educational and Psychological
Measurement, 54(3), 757-765. doi: 10.1177/0013164494054003022
Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). New York:
Guilford Press.
Kovacs, M. (2010). Children's Depression Inventory 2. Toronto, Ontario, Canada:Multi-Health
Systems.
MacCallum, R. C., Widaman, K. F., Zhang, S., & Hong, S. (1999). Sample size in factor analysis.
Psychological Methods, 4(1), 84-99. doi: 10.1037/1082-989x.4.1.84
Mejia, A., Filus, A., Calam, R., Morawska, A., & Sanders, M. R. (2014). Translation and validation of
the CAPES in Spanish: A brief instrument for assessing child maladjustment and parent
efficacy. Submitted for publication.
Mieloo, C., Raat, H., Oort, F. v., Bevaart, F., Vogel, I., Donker, M., & Jansen, W. (2012). Validity
and Reliability of the Strengths and Difficulties Questionnaire in 5–6 Year Olds: Differences by
Gender or by Parental Education? PLoS One, 7(5), e36805. doi: 10.1371/journal.pone.0036805
Morawska, A., & Sanders, M. R. (2006). Self-directed behavioural family intervention. International
Journal of Behavioral Consultation and Therapy, 2, 332-340.
Morawska, A., & Sanders, M. R. (2007). Concurrent predictors of dysfunctional parenting and
parental confidence: Implications for parenting interventions. Child: Care, Health &
Development, 33, 757-767.
Morawska, A., & Sanders, M. R. (2010). The Child Adjustment and Parent Efficacy Scale (CAPES).
Brisbane: Parenting and Family Support Centre.
Morawska, A., & Sanders, M. R. (2011). Disorders of Childhood. In E. Rieger (Ed.),Abnormal
Psychology: Leading Researcher Perspectives (2nd ed., pp. 403-447). Sydney: McGraw-Hill
Australia.
Morawska, A., Tometzki, H., & Sanders, M. R. (2012). An Evaluation of the Efficacy of a Triple PPositive Parenting Program Podcast Submitted for publication.
Myers, N., Ahn, S., & Jin, Y. (2011). Sample Size and Power Estimates for a Confirmatory Factor
Analytic Model in Exercise and Sport: A Monte Carlo Approach. [Article]. Research Quarterly
for Exercise & Sport, 82(3), 412.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory 3rd edition: McGrawHill.
Palmieri, P. A., & Smith, G. C. (2007). Examining the Structural Validity of the Strengths and
Difficulties Questionnaire (SDQ) in a U.S. Sample of Custodial Grandmothers. Psychological
Assessment, 19(2), 189-198. doi: 10.1037/1040-3590.19.2.189
Sanders, M. R. (2012). Development, Evaluation, and Multinational Dissemination of the Triple PPositive Parenting Program. Annual Review of Clinical Psychology, 8, 345-379. doi:
10.1146/annurev-clinpsy-032511-143104
Sanders, M. R., Markie-Dadds, C., Tully, L. A., & Bor, W. (2000). The Triple P- Positive Parenting
Program: A comparison of enhanced, standard, and self-directed behavioral family intervention
for parents of children with early onset conduct problems. Journal of Consulting and Clinical
Psychology, 68(4), 624-640. doi: 10.1037//002-006X.68.4.624
Sanders, M. R., & Morawska, A. (2010). Family Background Questionnaire.
Brisbane: Parenting
and Family Support Centre.
12
Australian Psychologist (Accepted 19/2/14).
13
Sanders, M. R., Ralph, A., Sofronoff, K., Gardiner, P., Thompson, R., Dwyer, S., & Bidwell, K.
(2008). Every Family: A population approach to reducing behavioural and emotional problems
in children making the transition to school. Journal of Primary Prevention, 3, 197-222. doi:
10.1007/s10935-008-0139-7
Sanders, M. R., & Woolley, M. L. (2005). The relationship between maternal self-efficacy and
parenting practices: Implications for parent training. Child: Care, Health & Development,
31(1), 65-73.
Smedje, H., Broman, J.-E., Hetta, J., & von Knorring, A.-L. (1999). Psychometric properties of a
Swedish version of the "Strengths and Difficulties Questionnaire". European Child &
Adolescent Psychiatry, 8(2), 63-70.
Spence, S. H. (1998). A measure of anxiety symptoms among children. Behaviour Research and
Therapy, 36, 545-566.
Stevens, J. (1992). Applied multivariate statistics for the social sciences: Lawrence Erlbaum.
Sumargi, A., Sofronoff, K., & Morawska, A. (2013). Understanding Parenting Practices
and Parents' Views of Parenting Programs: A Survey Among Indonesian Parents Residing in
Indonesia and Australia. In Press: Journal of Child and Family Studies Accepted 1/8/13.
Tsang, K. L. V., Wong, P. Y. H., & Lo, S. K. (2012). Assessing psychosocial well‐being of
adolescents: A systematic review of measuring instruments. Child: Care, Health and
Development, 38(5), 629-646. doi: 10.1111/j.1365-2214.2011.01355.x
Vimpani, G., Patton, G., & Hayes, A. (2002). The relevance of child and adolescent development for
outcomes in education, health and life success. In A. Sanson (Ed.), Children's health and
development: New research directions for Australia (pp. 14-37). Melbourne: Australian
Institute of Family Studies.
Webster-Stratton, C., & Reid, M. J. (2003). Treating conduct problems and strenghtening social and
emotional competence in young children: The dina dinosuar treatment program. Journal of
Emotional and Behavioral Disorders, 11, 130-143.
13
Australian Psychologist (Accepted 19/2/14).
14
Table 1.
Confirmatory Factor Analysis of the Factor Structure of CAPES Intensity.
Model
χ2
df
Δχ2
Δdf
CFI
SRMR
RMSEA
RMSEA
90% CI
Pre-test
A
701.71** 119
.742
.100
.128
.111 - .128
One-factor
B
408.31** 118 293.40**
1
.871
.075
.084
.076 - .093
Two-factor
Post-test
C
Parcel 5 & item 207.04** 89 201.27** 29 .936
.055
.062
.051 - .073
16 deleted
D
with correlated
196.40** 88
10.64**
1
.941
.055
.060
.049 - .071
error between
Parcels 2 & 12
E
with correlated
183.87** 87
12.53**
1
.947
.053
.057
.045 - .068
error between
Parcels 3 & 13
F
with correlated
174.88** 86
8.99**
1
.952
.052
.055
.043 - .066
error between
Parcels 9 & 11
Notes: df = degrees of freedom, CFI = comparative fit index, SRMR = standardized root
mean square residual, RMSEA = root mean square error of approximation, CI = confidence
interval. Models A to F are based on N = 346.
The chi square test is used to test the newly specified model with the previous one (only if the
models are nested, i.e., if one of the models could be obtained simply by fixing/ eliminating
parameters of the other model). This means that in the above Table Δχ2 test show the results
for comparisons of: the Model B with Model A; Model C with Model B; Model D with
Model C; Model E with Model D and Model F with Model E, respectively.
**p < .001
14
Australian Psychologist (Accepted 19/2/14).
15
Table 2.
Confirmatory Factor Analysis of the Factor Structure of CAPES Self-efficacy.
Model
χ2
df
Δχ2
Δdf
CFI
SRMR
RMSEA
RMSEA
90% CI
.103 - .138
A
176.58** 35
.950
.031
.120
B
with correlated
129.84** 34
46.74**
1
.966
.027
.100
.081 - .119
error between
Parcels 6 & 10
C
with correlated
101.98** 33
27.86**
1
.976
.023
.086
.068 - .106
error between
Parcels 2 & 7
D
with correlated
89.06**
32
12.92**
1
.980
.022
.080
.060 - .100
error between
Parcels 3 & 4
Notes: df = degrees of freedom, CFI = comparative fit index, SRMR = standardized root
mean square residual, RMSEA = root mean square error of approximation, CI = confidence
interval. Models A to D are based on N = 281.
The chi square test is used to test the newly specified model with the previous one (only if the
models are nested, i.e., if one of the models could be obtained simply by fixing/ eliminating
parameters of the other model). This means that in the above Table Δχ2 test show the results
for comparisons of: the Model B with Model A; Model C with Model B; Model D with
Model C.
**p < .001
15
Australian Psychologist (Accepted 19/2/14).
16
Figure 1
2-Factor Confirmatory Factor Analysis of the 27-item CAPES Intensity with 3 Error
Covariances and Standardized Estimates.
Item 5
.67
Item 17
Item 29
.79
Emotional
Maladjustment
.68
.36
Parcel 1
.81
-.18
Behaviour
Parcel 2
.73
-.22
Parcel 3
.81
-.17
Parcel 4
Parcel 6
.54
.76
.47
Parcel 7
.62
Parcel 8
.54
Parcel 9
.53
.45
Parcel 10
.68
Parcel 11
.56
Parcel 12
Parcel 13
16
χ2 (86) = 174.88, p < 0.001;
CFI = .952; SRMR = .052;
RMSEA = .055 (90% CI .043 - .066)
Australian Psychologist (Accepted 19/2/14).
17
Figure 2
1-Factor Confirmatory Factor Analysis of the 20-item CAPES Self-efficacy with 3 Error
Covariances and Standardized Estimates.
Parcel 1
.86
Parcel 2
.85
Confidence
Parcel 3
.87
.23
Parcel 4
.75
-.38
Parcel 5
.87
.84
Parcel 6
.85
Parcel 7
.89
.40
Parcel 8
Parcel 9
.89
.80
Parcel 10
17
χ2 (32) = 89.06, p < 0.001;
CFI = .980; SRMR = .022;
RMSEA = .080 (90% CI .060 - .100)
Australian Psychologist (Accepted 19/2/14).
18
Appendix A
Child Adjustment and Parent Efficacy Scale (CAPES)
Please read each statement and select a number 0, 1, 2 or 3 that indicates how true the statement
was of your child (aged 2-12) over the past four (4) weeks. Then, using the scale provided,
write down the number next to each item that best describes how confident you are that you can
successfully deal with your child’s behaviour, even if it is a behaviour that rarely occurs or does
not concern you.
There are no right or wrong answers. Do not spend too much time on any statement.
Example:
My child:
Gets upset or angry when they don’t get their own way
0
1
2
3
9
The rating scale is as follows:
0.
Not true of my child at all
1.
True of my child a little, or some of the time
2.
True of my child quite a lot, or a good part of the time
3.
True of my child very much, or most of the time
How true is this
of your child?
My child:
1.
ets upset or angry when they don’t get their own way
2.
efuses to do jobs around the house when asked
3.
orries
4.
oses their temper
5.
isbehaves at mealtimes
6.
rgues or fights with other children, brothers or sisters
7.
efuses to eat food made for them
8.
akes too long getting dressed
9.
urts me or others (e.g., hits, pushes, scratches, bites)
10.
nterrupts when I am speaking to others
11.
eems fearful and scared
12.
isbehaves at school or daycare
13.
18
Not at
all
A
little
Quite
a lot
Very
much
0
1
2
G3
0
1
2
R 3
0
1
2
W3
0
1
2
L 3
0
1
2
M3
0
1
2
A 3
0
1
2
R 3
0
1
2
T 3
0
1
2
H3
0
1
2
I 3
0
1
2
S 3
0
1
2
M3
0
1
2
H3
Rate your
confidence
1 = Certain I
can’t do it
10 = Certain I
can do it
Australian Psychologist (Accepted 19/2/14).
19
as trouble keeping busy without adult attention
14.
ells, shouts or screams
15.
hines or complains (whinges)
16.
cts defiant when asked to do something
17.
ries more than other children their age
18.
udely answers back to me
19.
eems unhappy or sad
20.
as trouble organising tasks and activities
0
1
2
Y 3
0
1
2
W3
0
1
2
A 3
0
1
2
C 3
0
1
2
R 3
0
1
2
S 3
0
1
2
H3
How true is this
of your child?
My child:
21.
an keep busy without constant adult attention
22.
ooperates at bedtime
23.
an do age appropriate tasks by themselves
24.
ollows rules and limits
25.
ets on well with family members
26.
s kind and helpful to others
27.
eems to feel good about themselves
28.
ets on well with other children
29.
alks about their views, ideas and needs appropriately
30.
oes what they are told to do by adults
19
Not at
all
A
little
Quite
a lot
Very
much
0
1
2
C 3
0
1
2
C 3
0
1
2
C 3
0
1
2
F 3
0
1
2
G3
0
1
2
I 3
0
1
2
S 3
0
1
2
G3
0
1
2
T 3
0
1
2
D3
Australian Psychologist (Accepted 19/2/14).
20
Appendix B
Table 3
Descriptive statistics for the items of CAPES Intensity Scale.
Item
Mean
SD
Range (min-max)
Item 1
2.62
.82
3 (1-4)
Item 2
2.03
.79
3 (1-4)
Item 3
2.21
.88
3 (1-4)
Item 4
2.02
.93
3 (1-4)
Item 5
1.86
.87
3 (1-4)
Item 6
2.30
.85
3 (1-4)
Item 7
1.73
.85
3 (1-4)
Item 8
1.97
.98
3 (1-4)
Item 9
1.46
.70
3 (1-4)
Item 10
2.46
.89
3 (1-4)
Item 11
1.55
.75
3 (1-4)
Item 12
1.53
.82
3 (1-4)
Item 13
1.70
.86
3 (1-4)
Item 14
2.02
.91
3 (1-4)
Item 15
2.20
.78
3 (1-4)
Item 16
1.97
.81
3 (1-4)
Item 17
1.41
.75
3 (1-4)
Item 18
1.92
.83
3 (1-4)
Item 19
1.55
.72
3 (1-4)
Item 20
1.81
.89
3 (1-4)
Item 21
3.14
.87
3 (1-4)
Item 22
3.06
.95
3 (1-4)
Item 23
3.65
.67
3 (1-4)
Item 24
3.12
.76
3 (1-4)
Item 25
3.42
.79
3 (1-4)
Item 26
3.24
.75
3 (1-4)
Item 27
3.17
.84
3 (1-4)
Item 28
3.34
.79
3 (1-4)
Item 29
3.10
.79
3 (1-4)
Item 30
3.25
.76
3 (1-4)
20
Skew
.29
.66
.57
.70
.74
.53
1.06
.70
1.51
.32
1.36
1.55
1.12
.65
.63
.71
1.93
.76
1.37
.90
-.63
-.64
-2.10
-.40
-1.23
-.64
-.72
-1.00
-.36
-.69
Kurtosis
-.73
.29
-.25
-.31
-.24
-.24
.50
-.55
1.87
-.69
1.59
1.68
.51
-.32
.27
.29
3.17
.17
1.94
-.03
-.55
-.64
4.20
-.61
.72
-.28
-.26
.30
-.82
-.15
Australian Psychologist (Accepted 19/2/14).
21
Table 4
Descriptive statistics for the items of CAPES Self-efficacy Scale.
Item
Mean
SD
Range (min-max)
Item 1
7.35
2.25
9 (1-10)
Item 2
7.82
2.23
9 (1-10)
Item 3
7.41
2.24
9 (1-10)
Item 4
7.70
2.38
9 (1-10)
Item 5
7.98
2.27
9 (1-10)
Item 6
7.33
2.23
9 (1-10)
Item 7
8.07
2.24
9 (1-10)
Item 8
8.12
2.18
9 (1-10)
Item 9
8.21
2.44
9 (1-10)
Item 10
7.65
2.10
9 (1-10)
Item 11
7.99
2.23
9 (1-10)
Item 12
8.20
2.24
9 (1-10)
Item 13
8.26
1.95
9 (1-10)
Item 14
7.66
2.35
9 (1-10)
Item 15
7.58
2.24
9 (1-10)
Item 16
7.70
2.23
9 (1-10)
Item 17
8.32
2.23
9 (1-10)
Item 18
7.73
2.39
9 (1-10)
Item 19
7.83
2.40
9 (1-10)
Item 20
7.95
2.06
9 (1-10)
21
Skew
-.79
-1.07
-.87
-1.10
-1.14
-.71
-1.21
-1.28
-1.47
-.95
-1.25
-1.46
-1.43
-.97
-.84
-.92
-1.55
-1.04
-1.13
-1.15
Kurtosis
-.01
.53
.16
.52
.51
-.24
.73
.90
1.29
.58
1.01
1.64
1.95
.14
-.16
-.02
1.75
.16
.45
1.07