Week 6: Reliability Flashcards
validity: design vs measurement
[statistical conclusion validity, internal validity, external validity] =/= [construct validity, measurement reliability]
6 indicators of construct validity
- face validity
- content validity
- predictive validity
- concurrent validity
- convergent validity
- discriminative validity
face validity
content of measure appears to reflect construct being measured (eg rating scale)
- based on judgement of instrument
content validity
- comparing content of measurement w content that defines construct
- questionnaire w good content validity should include questions on all components
predictive validity
- measurement to predict future behaviour/outcome
concurrent validity
- relationship between measurement and criteria at the same time (concurrently)
- important when trying to validate new instrument
- demonstrated when test correlates well w measure previously validated
convergent validity
measuring 2 theoretical constructs that are related in theory -> measurements should also be related
- extent to which scores are related to other measures
discriminative validity
- degree to which measurement is NOT related to other (not theoretically related) measurements
threats to construct validity
- inadequate description of construct
- inadequate measurement
- inadequate attention to IV levels
inadequate description of construct
- focus on limited number of components -> concluding that entire construct is evaluated
inadequate measurement
only one measurement/method to gather data
inadequate attention to IV levels
only one/two levels of multilevel variable are used
reliability
- replicability/consistency
internal reliability
- reliability of instrument chosen to evaluate DV in study
reliability coefficient
#; relationship between multiple test administration/multiple items - 0 - 1 scale (0.8 is good)