Research Methods Exam 12 June Thursday afternoon

Research Methods
Exam 12th June
Thursday afternoon
They ALWAYS
What are hypotheses?
contain 2
They are statements (NOT QUESTIONS!!!!)
variables
about the outcome of an experiment.
I.e. There is a link between being blonde and
being an awesome teacher
Or
There is a correlation between being in Miss
Baynes’ class and getting an A*!
What is a H1 and a H0?
• H1 – an alternate hypothesis which predicts a
difference or correlation. E.g. eating parmos
makes my bum bigger.
• H0 – a null hypothesis (null means no) which
predicts NO difference or correlation. E.g.
eating parmos will NOT make my bum bigger.
H1
H0
What are INDEPENDENT and
DEPENDENT variables
IV – the thing the psychologist
controls (THE CAUSE)
DV – the thing that is measured (THE
EFFECT)
E.g. If I did an experiment to see if
smelly students get better grades, my
IV would be the stench of the student
What type of variables are these?
1. There is a difference in how participants rate
their mood when in blue rooms and red
rooms.
2. Men are slower drivers than women
3. The amount of parmos I eat effects the size
of my butt.
IV
DV
Telling a joke
Age
Sex
amount remembered
attitude
amount they touch
Age
how they are portrayed
Sex
how much they forgive
Type of treatment
how effective it is
What are extraneous variables
and how can they be controlled?
They are variables that make the results
unreliable as they can have an effect on the DV.
E.g. In Haber and Levin, Joe might have not been able to tell how
far away the teddy was because it was foggy that day.
Standardisation is a way of controlled the
extraneous variables so that they remain
constant for all trials. E.g. Haber & Levin should test
everyone in the fog then!
What is a representative sample?
It’s when you take a sample of the TARGET
POPULATION (your target population could be all
teenagers in Brotton, or all ugly people in
Middlesbrough or all suicidal teachers) that is an
accurate reflection of the group.
E.g. within the suicidal teachers sample, does your
sample contain old and young teachers, male and
female, maths, english and science teachers, smelly
teachers, fat teachers, thin teachers etc?
What type of sampling would you
use?
Random – names out of a hat, everyone has
equal chance of being picked. Better chance
of a representative sample, unlikely to get
freak sample. Takes ages and people might
not want to take part.
Opportunity – available or convenient
participants. Quick and easy but usually not
representative.
Experiments!!
Repeated measures!
One group does two
different trials and you
compare the results.
Independent Groups
Two groups do different conditions
and you compare the results
CONDITION A
CONDITION A CONDITION B
+ Not as many participants
needed
+ No participant variables either!
- Order effect – bored, tired,
better!
CONDITION B
+ No order effect
-
Participant variables
Need more participants
Laboratory experiments
Takes part in a CONTROLLED
ENVIRONMENT
+ Easy to establish cause and effect if
environment is controlled, i.e. NO
EXTRANEOUS VARIABLES!!!
- Lacks ecological validity as not a
real life setting
- More DEMAND CHARACTERISTICS.
The environment sometimes gives
away what the experiment is about
and so people behave differently!
Field Experiments
An experiment carried out in a natural setting
+ High ecological validity as carried out in a real life
setting
- Less control over variables, so get extraneous
variables which effects reliability of results.
Ethical considerations
Informed consent – agrees to take part knowing
what the aim is
Right to withdraw – stopping taking part at any
time
Confidentiality – Protecting the identity (Don’t
tell anyone!!!)
Debriefing – check they’re okay before leaving!
WHY????
Why do we need ethics? No deception, no
distress,
no embarrassment, no harm!
Questionnaires
A.K.A a self report!
Open questions – e.g. what’s your
favourite colour? Dark white.
Closed questions – e.g. is your
favourite colour white? Yes/No
• Access thoughts and
feelings
• Same questions – patterns
and trends easily found
• Easy to administer to lots
of people quickly, lots of
data!! Facebook.
• LIARS!!!!! Social desirability
• Don’t understand
questions
• Closed questions – no
correct answer there
• Can’t explore individual
responses if you ask
everyone same questions
Interviews
Also a SELF REPORT
Structured – question are pre set
Unstructured – more like
conversation, q’s based on answers.
• Can access thoughts
and feelings
• Double checking and
clarification
• LIARS!!! Social
desirability
• Have to be able to
explain your own
feelings and
thoughts.
Observations
OVERT
COVERT
Participant observation
is when the observer
takes part with the
group being observed.
This is so they can
observe more closely.
Non participant
observation – they
just watch and don’t
join in. This is so
they don’t effect the
dynamic of the
group.
Pro’s and Con’s of observations!
•
•
•
•
Overt – more ethical than
covert, no deception
involved.
Covert – people act
naturally as they don’t
know they are being
watched
Participant observations –
the experimenter
experiences the situation
from the participants point
of view
Non participant – more
objective when standing
back and just watching and
not joining in
• Overt – observer effect
when you know you are
being watched
• Covert – unethical
• Participant – effects
dynamics of group
• Non participant – miss
details
• OBSERVER BIAS – open to
interpretations. OVER
COME WITH INTER RATER
RELIABILITY (need more
than one observer)
Types of studies
CASE STUDY
Small sample, in depth analysis using
interviews, observations, records e.g.
criminal, school etc.
CORRELATION Collect 2 sets of data and look for a link between them.
Positive correlations – one thing increases so does the
STUDY
other. Negative correlation – one goes up, one goes
down!
LONGITUDINAL Long period of time, e.g. Mednick’s
adoption study
STUDY
CROSS
When you compare different groups, e.g.
SECTIONAL Piaget compared different age groups of
children
STUDY
Types of data
Quantative – NUMERICAL DATA
Qualitative - DESCRIPTIVE DATA
Number of words you can remember Quantative
QUALITATIVE
Diary entries of a serial killer
The average number of times someone stops to help a stranger
The different ways a teacher treats boys and girls
A tape recording of an interview with David Reimer (twin boy)
How long it takes to recognise a face
The mean, median or mode
Mean – the average
Median – middle number
Mode – most frequently occuring
Limitations
Reliability – more than one observer (inter rater
reliability). Make sure the data is
replicable, if you do it again will you
find the same results?
Validity
- Do the findings reflect the truth? Did
the questionnaire measure what it was
meant to measure? Did the study find a
real answer or did it have to many
problems to be valid?
Types of bias
Gender bias – when one gender is favoured over
another in the findings, doesn’t represent both
sexes – can’t be applied to both sexes
Cultural bias – one SOCIAL GROUP is favoured
over another in the findings, therefore findings
cannot be applied to a wider population
Experimenter bias – When the experimenter
interprets the results to support a theory.
Alternative evidence is overlooked or ignored.