validity

define validity/internal validity

is the extent to which a test measures what it claims to measure

why isn’t having consstency enough?

a measure might give the same result everytime, but if it isn’t measuring the right thing, results are meaningless

define external validity

whether the findings can be generalised beyond the research setting to other people, situations, or times

what affects internal validity

• researcher effects

• participant variables

• poor controls in places

• social desireability

what impacts external validity?

• biased sample

• very artificial conditions

which time of validity compares the current test to a previously validated test?

concurrent validity

define face validity

whether a test looks like its measuring what its intention is

how do you improve validity in questionnaires?

• reassure participants their answer is confidential

what’s a lie scale? and what does it do

• to indentify inconsistency or socially desireable responses (asking the same questions in different ways)

how to improve validity in observations

• state clear behavioural categories

how to improve validity in an experiment

• use a control group (check if DV is affected by IV)

• standardise procedures to reduce differences

• use blind procedures to reduce researcher effects + reduces demand charcteristics

what is confirmation bias?

is the tendency to search for, interpret and remember information that supports our existing beliefs, while ignoring information that contradicts them

whats it mean when variables aren't clearly defined or measured, meaning the study may not be measuring what it intends to measure

poor operationalisation

define confounding variables

an unwanted variable changes alongside the IV, meaning we cannot be certain the IV caused the results

what are the three types of external validity?

• ecological

• population

• temporal

define ecological validity

whether findings can be generalised to real-life settings

which type of validity explores whether findings can be generalised to different time periods?

temporal

define population validity

whether findings can be generalised to other people or populations