Validity Flashcards
internal validity is all about
control
internal validity is about what happens
within the study
to deal with low internal validity we need to
control extraneous/confounding variables
three participant variables
1) repeated measures/matched pairs design
2) random allocation to conditions
3) large sample
two situational variables
1) standardized procedures
2) controlled environment
two researcher bias
1) double blind
2) independent researcher
four demand characteristics
1) deception
2) single blind
3) filler tasks/questions
4) high realism
two order effects
1) counterbalancing
2) independent/matched measures design
problem: flawed behaviours categories
solutions:
1) pilot study
2) give the observers training and practice and time to discuss and ask questions
problem: presence of an observer
solution: covert observation
problem: observer bias
solutions:
1) more than one observer and correlate their data
2) if data collected is similar then there is unlikely to have been bias
problem: difficulty in recording accurate data
solutions:
1) use recording equipment
2) therefore observations can be replayed and analysed
what is the key problem in gathering accurate, valid data from self-report
social desirability bias, bad memory etc.
problem: no appropriate option to choose as an answer, so participants can’t put what they feel
solution: use open ended questions
problem: some people don’t know what they think or feel
solution: give p’s some options for possible answers
problem: giving socially acceptable answers rather than honest ones (social desirability bias)
solutions:
1) add a lie scale (this is a set of questions that determine if the participant is lying)
2) make the self-report anonymous and/or online, tell them answers are confidential
problem: leading questions/ambiguous (unclear) questions
solution: change or remove problematic questions
problem: interviewer bias - the interviewers’ expectations, beliefs and prejudices may influence the answers given by the p
solution: use a different research method e.g. observation, experiment
problem: interviewer effects - something about the characteristics of the interviewer influences the answers the participants give, e.g. age, gender, attractiveness
solution: use a questionnaire rather than an interview
external validity is about
generalizing
external validity is about what happens
to the findings after the study
improving external validity - ecological validity
change the location/setting to something more representative of every day life and natural environment. everyday life e.g. field - mundane realism
improving external validity - population validity
use a more representative sampling method e.g. stratified or random good especially if population is large - large sample
improving external validity - historical (temporal) validity
replicate the study in the current day to see if results are the same