convergent and divergent validity Flashcards
what is convergent and divergent validity?
reflects the degree to which test scores have the correct patterns of associations with other variables
what does the conceptual foundation of a construct include?
the connections between the construct and a variety of other psychological constructs
what are the interconnections between a construct and other related ones called?
nomological network
what is predictive validity?
where a particular correlation between test scores and a specific criterion are make or brake
what does it mean if the validity coefficient is large?
then we have more confidence in using the test for its intended purpose
what is validity generalisation?
a process of evaluating a tests validity coefficients across a large set of studies
what do validity generalisation studies evaluate?
the predictive utility of a test’s scores across a range of settings, times and siuations
what are the three questions that validity generalisation studies address?
- the average level of predictive validity across studies
- the degree of variability associated with the validity coefficients
- identify sources of systematic variability in the coefficients
what is a nomological network based on?
a large number of coefficients across a wide range of criteria
once the coefficients are estimated what happens?
they are placed on a table for visual inspection then a subjective judgement is made about the pattern of coefficients with respect to whether a judgement construct validity is meritied
what do multitrait multimethod matricies try to overcome?
the fact that a correlation between two scores may conflate two sources of variance
what are the two scores of variance that MTMM’s deal with?
trait variance (good stuff) method variance (bad stuff)
what do we want to see in terms of trait and method variance?
hope to see larger correlations between the same traits, in comparison to correlations between scores based on just the same method
do we want a lot of shared trait variance?
yes, particularly identical traits measured using different methods
what does quantifying construct validity involve?
requires researchers to predict the magnitude of the correlation between their measure of interested and their selected criterion. then, the correlations between their measure of interest and these selected criteria are estimated. finally, the correlation between the predicted and estimated correlations is estimated.