Research Methods and Techniques Flashcards
experiment, observation, self report, correlation
IV
what the researcher wants to compare/manipulate/change
DV
what the researcher measures
extraneous variables
variables that could affect the DV if not controlled
confounding variables
variables that could affect the DV strongly enough to influence the results
lab experiment
- experiments conducted in an artificial environment
- researcher manipulates the IV to measure its affect on the DV
- controls extraneous variables which could influence the results
strengths of lab experiments
- high level of control so affect of EV are minimised
- increased replicability
- shows cause and effect
weaknesses of lab experiments
- low ecological validity
- prone to demand characteristics
- ethical concerns as deception is often used
Field experiment
- conducted in natural environments
- little control over extraneous variables
- IV is manipulated so cause and effect can be inferred
strengths of field experiments
- high ecological validity
- demand characteristics can be minimised
weaknesses of field experiments
- low control over variables
- difficult to replicate
- difficult to record data
- ethical concerns: lack of consent, deception, stress, no debriefing
quasi experiments
- naturally occurring IV
- usually conducted in a lab
strengths of quasi experiments
- naturally occurring IV
- highly controlled so the effect of EV is minimised
- can show cause and effect
weaknesses of quasi experiments
- low ecological validity
- not easy to replicate
- prone to demand characteristics
- ethical concerns: deception is often used & stress may occur
participant observations
- researcher becomes part of the group whose behaviour is being observed
- can be overt or covert
non-participant observations
- researcher records behaviour without being involved in the situation being observed
- can be overt or covert
structured observation
- researcher has a coding frame for recording behaviour
- can be participant or non participant, overt or covert
unstructured observation
- researcher records all behaviour that is relevant to the study
- can be participant or non participant, overt or covert
controlled observation
occur in an artificial setting and it is more likely the participant will be aware they are being observed
naturalistic observation
occur in a more natural setting so have lower-levels of control but higher levels of realism
overt observation
participants are aware they are being observed
covert observation
participants are unaware they are being observed
strengths of observations
- observations give a different take on behaviour, as what people say they will do and actually do are often different
- able to capture spontaneous and unexpected behaviour
- high in ecological validity
- coding frames make recording behaviour easy
weaknesses of observations
- observer bias may occur (may see what they want to see)
- poorly designed coding frames reduce reliability and inter-rater reliability
- participants may show demand characteristics or socially desirable behaviour
- observations cannot provide information about what people think or feel
self report
interviews or questionnaires to collect data
questionnaires
completed by the participants or a set of questions read to the participants by the researcher
structured interviews
- asking predominantly closed questions in a fixed order
- scripted in a standardised order and every participant is asked the same Q in the same order
semi structured interviews
- fixed list of open/closed questions
- researcher can introduce questions if required to clarify/expand on a topic
unstructured interview
- starts with standard questions asked to every participant
- questions then depend on participants answers
- interviewer may have a list of topics to cover
strengths of self reports
- large amounts of standardised data can be gathered from a large number of participants quickly/conveniently so results are more generalisable
- questionnaires/structured interviews are highly replicable so trends/similarities/differences in behaviour can be identified
- unstructured interviews allow in depth, qualitative data to be gathered
weaknesses of self reports
- questionnaires/structured interviews lack flexibility so participants cannot expand on answers
- unstructured interviews cannot be replicated
- responses are often influenced by demand characteristics/social desirability
what is correlation used for
- to establish whether there is a relationship between 2 variables
- does not involve any manipulation
3 types of correlation
- positive
- negative
- no/zero
positive correlation
- as scores on one variable increase, so does the other
- does not show cause and effect but does show they are positively related
negative correlation
- As the scores on one variable increases, the other decreases
- does not indicate cause and effect but does show they are negatively related
strengths of correlations
- uses quantifiable measures of each variable
- gives information about the relationship between the variables measured
- strong negative/positive correlations form the basis for further researcher to establish cause and effect
weaknesses of correlations
- cause and effect cannot be established
- findings can be misleading as they can imply cause and effect relationships that don’t exist
Strengths of participant observations
- Gives observer insight into participant emotions and motives
- participants may behave more naturally if they’re unaware if the observers
- people may reveal more if they are unaware of the observer
Weaknesses of participant observations
- participants are aware they are being observed (but not by whom) so may not reflect normal behaviour
- observer may become subjective if they get involved with social group
- if participant observer is hidden, it raises ethical and practical issues as participants can’t give informed consent
Strengths of non-participant observations
- if observer is overt, data recording systems make more accurate and detailed records
- observers remain objective from the social situation
Weaknesses of non-participant observations
- ethical questions raised if participants are unaware they’re being observed (right to withdraw)
-if participants are aware of the observer, thus may impact behaviours displayed
Strengths of overt observations
- more ethical way to observe behaviour as the participants are aware of the observer’s presence
Weaknesses of overt observations
- if the participants are aware of being observed, this can impact their behaviour
Strengths of covert observations
-participants will act more naturally if they’re unaware they’re being observed
- if covert and non-participant then data recording devices can increase accuracy
Weaknesses of covert observations
- if observer is identified or suspected then validity is compromised
- ethical questions raised, consent etc
- if the observer is a participant observer, becomes harder to record data accurately without being detected
Strengths of structured observations
- opertational definitions can be developed in a pilot study increasing validity
- operationally defined behaviour categories can be agreed between observers (more reliable)
- practising the use of data collection techniques improves inter-rater reliability
- photographic and video data collection allows for reanalysis of data
Weaknesses of structured observations
- simple definitions may not convey sufficient meaning
- total numbers of behaviours in a category can be meaningless without context, lowers validity
- predetermined behaviours may become limiting if new behaviours occur during the study
Strengths of unstructured observations
- without any predetermined behaviours, anything relevant can be recorded which provides richer data
- detailed descriptions rather than simple catagory totals means you get a more complete picture
Weaknesses of unstructured observations
- attempting to record everything means certain aspects may be missed or ignored
- some data may be irrelevant from key features
- without operational definitions, recording may be inconsistent or subjective
What is counterbalancing
- ABBA design
- sample is split into 2 groups
- group one does A then B, group 2 does B then A
Why is counterbalancing used
- eliminates order effects
- order effects still occur in both groups, but they cancel out as they occur equally in both groups
Time sampling
behaviours are recorded at predetermined intervals such as every 5 minutes
Instantaneous scan sampling
- behaviour is recorded at the start of the time interval
- e.g., every 10 seconds a researcher may record the action performed by the subject and any other actions being ignored
predominant activity sampling
at a determined interval e.g., every 10 seconds, the most common behaviour is recorded
One-zero time sampling
at timed intervals, coding frame is used to record whether a behaviour did or didn’t happen
Event sampling
recording pre-established behaviours each time they occur
