Assessing reliability and validity Flashcards
what is the test re-test method for assessing reliability?
-carry out the test once.
-carry out the test again on the SAME people at a LATER date.
-CORRELATE to check if they are similar.
correlation of +0.8=reliable
what is the inter-observer method for assessing reliability?
-have a SECOND researcher carry out the research using the SAME behavioural categories
-they both observe the SAME thing, but SEPERATELY.
-CORRELATE +0.8=reliable
how do you improve the reliabilty of a quesionnaire?
remove or REWRITE some questions if they are ambiguous, replace open questions with closed to reduce ambiguity.
how do you improve the reliability of an interview?
use the SAME interviewer each time.
if not, TRAIN all interviewers to AVOID leading questions, and use STRUCTURE to have more control over questions.
how do you improve the reliability of an experiment?
use a LAB experiment so that it has CONTROL and is consistent, so allows method to be replicable.
make sure it is STANDARDISED.
how do you improve the reliability of an observatioin?
make sure behavioural categories are OPERATIONALISED (easily measurable) and don’t overlap. TRAIN observers.
what is internal validity?
whether or not we are measuring what we are supposed to be measuring,
what is face validity?
it LOOKS like it’s measuring what it’s supposed to be.
what is concurrent validitiy:
the results from this measure are CLOSE to the results from another ALREADY established test.
what is ecological validity?
the extent to which the findings can be generalised to other settings and situations. (strengthens its validity)
what is temporal validity?
the extent to which findings can be generalised to other HISTORICAL times or eras.
how do you assess face validity?
researcher will look at test and see if i APPEAR to be measuring what it’s supposed o be.
how do you assess concurrent validity?
give the test to participants and collect results.
test them on an already ESTABLISHED test and collect results.
CORRELATE the scores (+0.8)
how do you improve the validity of an experiment?
-CONTROL extran variables to stop them being confounding.
e.g.
use control group to see if changes are affected by the IV.
standardise procedures or use double blind procedure to reduce investigator effects.
how do you improve the validity of a questionnaire?
include a LIE scale to see how consistent responses are to reduce social desirability bias.
make it anonymous
