research methods Flashcards
4 types of study
lab, field, quasi and natural
empiricism
evidence gained through observation and experiment
independent variable
is what you change
dependent variable
is what you measure
falsifiable
hypothesis you can test and prove right or wrong
advantages of lab experiments
- high control over variables
- high internal validity
- easier to replicate (reliable)
disadvantages of lab experiments
- lack generalisability
- low ecological validity
- demand characteristics
advantage of field experiments
- more valid and authentic
- high external validity
disadvantages of field experiments
- loss of control over variables
- ethical issues (consent)
advantages of natural experiments
- provide new opportunities for research
- high external validity
disadvantages of natural experiments
- very rarely have the opportunity
- no generalisibility
independent measures
when a participant is tested once in one condition of the experiment
advantages of independent measures
- avoids order effect
- reduce demand characteristics
disadvantages of independent measures
- more people are needed
- variables between participants
repeated measures
the same participants are allocated to all groups so get tested in all conditions
advantages of repeated measures
- no individual differences
- less participants needed
disadvantages of repeated measures
- order effects
- different tests and materials are needed
matched pairs
participants take part in one condition then are matched with someone similar to do the other conditions
advantages of matched pairs
- no order effects
- same tests and materials can be used
disadvantages of matched pairs
- some participant variables
- matching is hard and time consuming
random sampling
simple, systematic, stratified
non random sampling
opportunity, quota, snowball
simple random
randomly selecting people from a list of names
systematic random
numbering participants and picking them at set intervals
stratified random
dividing the group into smaller sections which reflect the proportions of the target population
opportunity sampling
selecting people who are willing to take part
volunteer sampling
participants select themselves to be involved
+/- simple random
+ no researcher bias
- not representative
+/- systematic sampling
+ no researcher bias
+ quite representative
+/- stratified
+ representative
- could have researcher bias
+/- opportunity
+ less time and money
- researcher bias
+/- volunteer
+ minimal effort
- volunteer bias
ethical issues
Deception Protection from harm Right to withdraw Debrief Informed consent Confidentiality
pilot studies
are test runs of the research to help the researcher identify any issues
overt non participant observation
openly observing peoples behaviour with consent
covert non participant observation
observing people who are unaware they’re being observed
overt participant observation
being involved with the participant while observing them with consent
covert participant observation
being involved with the participant but not telling them you’re observing them
controlled observation
certain aspects of the observation are controlled (Strange Situation)
event sampling
counting the number of times a particular behaviour occurs
time sampling
recording behaviour within a pre-established time frame
advantages of structured interview
- standardised data
- easy to compare respondents
- easy to replicate
- large numbers
- quick and cheap
disadvantages of structured interview
- lack validity
- lack depth
- lack flexibility
- the same question could be interpreted differently
advantages of unstructured interview
- validity and depth
- better understanding respondents
- leads to new ideas
- reduced interviewer effect
disadvantages of unstructured interview
- less standardised
- less reliable
- more time and money
- avoid cherry picking information to fit hypothesis
advantages of questionnaires
- collect information from a lot of people
- can use all types of questions
- identifies correlations
- representative and reliable
disadvantages of questionnaires
- closed and graded limit response
- may lack validity ( demand characteristics)
likert scale
respondent indicates their agreement with a statement on a scale of usually 5 points
normal distribution
a bell shaped curve on a graph
positive skew
most of the distribution is towards the left of the graph
negative skew
most of the distribution is towards the right of the graph
directional hypotheses
the researcher makes clear the different that is anticipated between 2 conditions
non directional hypotheses
simply states there’s a difference between conditions or groups, but the nature of it is not specified
extraneous variables
other variables that might potentially interfere with the independent or dependent variable
randomisation
the use of chance to reduce the researchers influence in the investigation
quasi experiment
have an independent variable based on existing difference between people (age, gender)
+ controlled lab experiments
- confounding variables
single-blind procedure
the participants are unaware of the experiment aims
double-blind procedure
the researcher and participant are both unaware of the experiment aims
behavioural categories
a target behaviour is broken up into components that are measurable
co-variable
the variables investigated within a correlation
positive correlation
as one co-variable increases, so does the other
negative correlation
as one co-variable increases, the other one decreases
measures of central tendency
are averages which give information about the most typical values in a set of data
types of graph
bar, scattergram, histogram
percentages
certain number / total number x 100
decimal
remove the % sign and move the decimal point 2 places to the left
fraction
the number of decimal places is the number of zeros to divide by
the critical value
- the significance level (0.05)
- the number of participants
- directional or non directional hypotheses
sign test 1
subtract the 2nd column from the 1st column to find the sign of difference (+/-)
sign test 2
calculate the total number of + and - and the less frequent one is S
sign test 3
the calculated value of S must be less than/equal to the critical value
