Research Methods: Validity Flashcards
Define Temporal Validity.
A type of external validity. Research has temporal validity when the procedure and setting of the research approximates the real world, meaning the findings can be generalised to the real world.
Define Ecological validity.
A type of external validity. Research has ecological validity when the procedure and setting of the research approximates the real world meaning the findings can be generalised to the real world.
Define Face Validity.
A simple way of assessing validity. We can assess if research has face validity by asking ourselves whether the research appears to measure what it claims to. If it does than it has face validity.
Define Concurrent validity.
Another method for assessing validity. Research that has concurrent validity will have similar findings to those from another well established test of the thing being measured.
How do you test for concurrent validity?
- Find a well established test/measure of the thing being measured by the research in question.
- Compare the performance of a sample of participants on both measures.
- If there is a correlation coefficient of 0.8 or above between the results on the two tests, this indicates close agreement and concurrent validity. A lower correlation would suggest the two tests are measuring different things.
How do you improve internal validity in Questionnaires and Interviews?
> Face Validity : Changing the direction of responses along rating scale (e.g., so
‘Agree’ isn’t always on the right)
Concurrent : Use Lie scale (one question is asked in two different ways (‘I never regret my actions’ and ‘Sometimes I feel bad about things I’ve done.’ You cannot answer yes to both of these and be truthful.)
Assure confidentially by making sure data is anonymous
How could this improve internal validity in questionnaires and interviews?
> Face validity controls for or response bias, which is an EV that can
undermine the validity of a questionnaire.
Concurrent validity controls for social desirability bias, which is an EV which can undermine the validity of a questionnaire.
It encourages participants to give truthful responses as
they will be less concerned about their responses
becoming public