Research Methods AO1 Flashcards
What is the aim of an experiment?
Purpose of the investigation
What are hypotheses?
Testable, operationalised statements
What are directional and non directional hypotheses?
Identifying a difference/correlation or not, choice depends on previous theory or research
What are extraneous variables?
Nuisance variables but randomly distributed, can affect DV if not controlled
What are confounding variables?
Vary systematically with IV, can’t tell if change in DV is due to IV or CV
What are demand characteristics?
Participants second guess study aims, alter behaviour
What are investigator effects?
Influence of researcher on DV/design decisions
What is randomisation?
Chance methods to reduce researcher’s bias
What is standardisation?
Ensuring all participants have the same experience
What is the independent variable?
Aspect of experiment that is manipulated/changes naturally so effect on DV can be measured
What is the dependent variable?
Variable that is measured by researcher, any effect should be caused by change in IV
What is operationalisation?
Defining variables so they can be measured
What are independent groups?
Participants in each condition of experiment are different
What are repeated measures?
All participants take part in all conditions
What are matched pairs?
Similar participants paired on participant variables, allocated to condition A or B
Evaluate independent groups
Participant variables not controlled (use random allocation)
Less economical
No order effects
Evaluate repeated measures
Order effects (use counterbalancing)
Demand characteristics
Participant variables controlled
More economical
Evaluate matched pairs
No order effects
Cannot match participants exactly
Time consuming
What are lab experiments?
IV manipulated in controlled setting, participants go to researcher
What are field experiments?
IV manipulated in natural setting, researcher goes to participant
What are natural experiments?
IV changed naturally. DV/setting may be natural or in a lab
What are quasi experiments?
IV based on an existing difference between people, effect on DV recorded
Evaluate lab experiments
High internal validity, control of CV/EV, cause and effect shown
Replication more possible, support for findings
Low external validity, mundane realism, artificial tasks, ungeneralisable
Low internal validity, demand characteristics
Evaluate field experiments
High external validity, more authentic, realism
Lower internal validity, less control
Ethical issues, consent not possible
Evaluate natural experiments
Only option for practical/ethical reasons
High external validity, real world problems
Limited opportunities
No random allocations
Low realism in a lab
No manipulation of IV, can’t claim cause and effect
Evaluate quasi experiments
If in lab, same issues
No random allocation
No manipulation of IV, can’t claim cause and effect
What is random sampling?
Equal chance of selection, lottery
What is systematic sampling?
Selecting every nth person from list
What is stratified sampling?
Sample reflects proportion of people in different population strata
What is opportunity sampling?
Choosing whoever is available
What is volunteer sampling?
Participants self select (through adverts for example)
Evaluate random sampling
Potentially unbiased
Control CV/EV
Time consuming
May not work
Evaluate systematic sampling
Objective method
Time consuming
Those selected may refuse
Evaluate stratified sampling
Representative
Cannot account for all subgroups
Evaluate opportunity sampling
Convenient
Unrepresentitive
Evaluate volunteer sampling
Easy
Participants engaged
Volunteer bias
Responsive to cues
What are ethical issues?
Conflict of rights of participants and aims of research
What is informed consent?
Advise participant what is involved, reveals research aims
Signed consent form, presumptive, prior general, retrospective
What is deception?
Misleading participants, withholding information
OK if not distressing
Debriefing, right to withdraw/withhold data, counselling
What is protection from harm?
Psychological/physical risk should be ‘normal’
What is privacy and confidentiality?
Right to control and protect personal data
Use numbers, not names, data not shared with other researchers
What are pilot studies?
Check procedures and techniques, make changes
What is single-blind?
Participants aren’t aware of aims/conditions until end
What is double-blind?
Neither participants nor individual conducting research know the aim/condition
What are control groups/conditions?
Used as a comparison or baseline, no variables manipulated
What are naturalistic observations?
Behaviour observed where it would normally occur
No control over variables
What are controlled observations?
Some control over environment, such as manipulation of variables
What are covert/overt observations?
Observing without or with participants knowledge
What are participant/non-participant observations?
Join group or remain outsider
Evaluate observations
Capture what people do
Observer bias
No casual relationships shown
Evaluate naturalistic observations
Low internal validity, control difficult
High external validity, everyday life
Evaluate controlled observations
High internal validity, extraneous variables controlled
Low external validity, except if covert
Evaluate covert and overt observations
Covert: low demand characteristics, ethically questionable
Overt: behaviour may be affected, more ethical
Evaluate participant and non participant observation
Participant: increased external validity, lose objectivity
Non-participant: more objectivity as increased internal validity, less insight
What are ways of recording data in observational design?
Record everything (unstructured) or categories (structured)
What are behavioural categories in observational design?
Target behaviours broken down into observable components
What are sampling methods in observation design?
Continuous
Event sampling: count events
Time sampling: count at timed intervals
Evaluate structured vs unstructured observational designs
Structured: numerical, easier to analyse
Unstructured: may just be eye catching information, qualitative data harder to analyse, observer bias
Evaluate behavioural categories in observational design
Must be observable
Avoid dustbin category, all possible forms of behaviour listed
No overlap
Evaluate sampling methods in observational design
Event: useful for infrequent behaviour, misses complexity
Time: less effort but may not represent whole behaviour
What are questionnaires?
Pre-set list
Can use to measure DV in experiments
What are closed and open questions?
Fixed choice or not, quantitative data or not
Evaluate questionnaires
Distribute to many people
Fixed choice
Easy to analyse
Social desirability response biasE
Evaluate closed and open questions
Quantitative or qualitative data, effects ease of analysis
What are structured interviews?
Pre-set questions, fixed order, face-to-face
What are unstructured interviews?
No formula, general topic
Questions based on responses
What are semi-structured interviews?
Some pre-set questions with follow-ups
Evaluate structured interviews
Easy to replicate
Interviewer can’t elaborate/explain
Evaluate unstructured intervuews
Flexible
Increased interviewer bias
Analysis more difficult
Social desirability bias reduced by rapport
What designs can be used in questionnaires
Likert scale, rating scale, fixed choice
How can interviews be designed?
Standardised schedule, avoids interviewer bias
Comfortable setting for rapport
Ethical issues
How should good questions be made?
Don’t overuse jargon, not too technical
Replace loaded phrases with neutral ones, don’t use emotive language or leading questions
Ask one question only, no double barrelled questions or doubled negatives
What are the three types of correlation?
Positive, negative and zero
What is the difference between correlations and experiments?
No manipulation of variables, no cause and effect
Evaluate correlations
Useful starting point
Quick
Economical
Secondary data
Can’t demonstrate cause and effect
Intervening variables
May be misinterpreted
What is qualitative data?
Written, non numerical description of participants thoughts, feelings, etc
What is quantitative data?
Expressed numerically rather than words
Evaluate qualitative data
Rich in detail
Greater external validity
Difficult to analyse
May be subjective
Evaluate quantitative data
Easy to analyse (graphs)
Less biased
Narrower in meaning
What is primary data?
Collected first hand for purpose of investigation
What is secondary data?
Collected by someone other than researcher
Evaluate primary data
Fits the job
Targets relevant information
Requires time
Evaluate secondary data
Inexpensive
Easy to access
Variation in quality, may be outdated
What is a meta analysis?
Type of secondary data, statistical analysis of larger number of studies
Produces effect size
Evaluate meta analysis
Large sample
High validity
Publication bias/file drawer problem
How to calculate mean?
Add up, divide by number
What is the median?
The middle value
What is the mode?
Most frequently occurring value
Evaluate the mean
Most sensitive and representative measure of central tendency
Easily distorted by extreme values
Evaluate the median
Less affected by extremes, not sensitive
Evaluate the mode
Relevant to categorical data
Crude, unrepresentative
How to calculate the range?
Subtract the lowest from the highest (add 1)
What is standard deviation?
How much scores (on average) deviate from mean
Evaluate the range
East to calculate
Unrepresentative if there are extremes
Evaluate standard deviation
More precise than range
Distorted by extreme values
How to conduct the sign test?
Convert to nominal data
Add up pluses, add up minuses
S = less frequent sign
Compare calculated S value with critical value, if S <= critical value, difference is significant
What are the aims of peer review?
Allocate funding
Validate quality of research
Suggest amendments/improvements
Evaluate peer review
Anonymity, may permit unjustified criticisms by rivals
Publication bias, file drawer problem, creates false impression of current knowledge
Burying groundbreaking research, maintains status quo
Give examples of psychological research and the economy
Attachment research: role of father, equal care from both parents means more effective contribution to economy
Mental health: absenteeism due to depression has a cost, psychological research reduced mental disorder (new drugs, CBT)
What are correlations?
Relationship between two continuous co-variables
What are correlation coefficients?
Represents strength and direction of relationship
How to interpret what co-efficitents mean?
The closer to -1 or +1, the stronger the relationship
Sign tells direction
What are case studies?
Detailed analysis of unusual individual or event, may also be ‘typical’ behaviours
What are the characteristics of a case study?
May involve case history
Qualitative (interviews) and quantitative (psychological tests) data
Tend to be longitudinal
Evaluate case studies
Insight into unusual cases
Generate hypotheses for future studies
Generalisation from small samples is a problem
Conclusions based on subjective interpretation of researcher
Subjective data from participants
What is content analysis?
A form of observation where communication is studied indirectly
How is coding and quantitative data used in content analysis
Data categorised into meaningful units, analysed by counting words etc
How is thematic analysis and qualitative data used in content analysis
Recurrent ideas that keep appearing in communication identified and described
Evaluate content analysis
Fewer ethical issues than case studies
High external validity
Flexible approach, can be adapted
Information may be studied out of context
May be subjective
Reflexivity aims to address issue of bias
What is reliability?
Any measurement should produce the same result unless the thing it is measuring has changed
What is test-retest?
The same test administered to the same person on different occasions, results compared
What is inter-observer reliability?
Observers compare data in a pilot study or at end of actual study to make sure behavioural categories are consistently applied
What is the null hypothesis?
States no difference between conditions
Statistical tests determine whether this should be accepted or rejected
What are levels of significance?
The point at which researcher can accept the alternative hypothesis (usually 5% in psychology)
How should statistical tables be used in psychology?
Calculated and critical values, calculated must be compared with critical to determine significance
One tailed or two tailed test required? What is n or df value? Which level of significance is required?
More stringent LoS should be used when human cost or one off study
What is a type 1 error?
Incorrect rejection of a true null hypothesis
More likely if significant level too lenient
Optimistic error
What is a type 2 error?
Incorrect acceptence of false null hypothesis
More likely if significance level too stringent
Pessimistic error
What is Mann-Whitney U test?
Non-parametric test
Difference between two sets of data
Unrelated design
Data at least ordinal level
What is Wilcoxon T test?
Non parametric test
Difference between two sets of data
Related design
Data at least ordinal level
What is the unrelated t-test?
Parametric test
Difference between two sets of data
Unrelated design
Data at interval level
What is a parametric test?
Used with interval level data, normal distribution expected and satisfied homogenity of variance (sd squared)
Data drawn from population with expected normal distribution and both sets have homogeneity of variance
What is a related t-test?
Parametric test
Test of difference between two sets of data
Related design
Data at interval level
What is Spearman’s rho?
Test of correlation between co-variables
Data at least ordinal level
What is Pearson’s r?
Test of correlation between co variables
Data at interval level
Parametric test
What is chi squared?
Test of difference between two sets of data or association between co-variables
Data is independent
Nominal data
What is rule of R?
Tests with R in their name are those where calculated value must be equal to or more than critical value
What is statistical testing used for?
Determine whether to accept or reject null hypothesis
How to decide on stats test?
- Difference or correlation? Correlation includes tests of association
- Experimental design? Related or unrelated?
- Level of measurement? Nominal, ordinal or interval?
What is nominal data?
Data represented in form of categories
What is ordinal data?
Ordered data, unequal intervals, can be placed in rank order, such as using a rating scale
What is interval data?
Based on numerical and public scales of measurements, units of equal size, like temperature
