Validity Flashcards
Fall 2018 Midterm
Describe Cronbach and Meel’s (1955) trintarian view of validity.
Criterion (ability of item to predict criterion), content (measurement represents all facets of given construct), and construct (measurement operation measures what you want it to) validity.
What is convergent and discriminant validity?
Convergent validity- measures differ in method but give the same answers.
Discriminant validity- when measuring different things on the same measure, correlation is not high.
What is the MTMM?
Multitract Multimethod Matrix. Describes every measure as a trait-method unit.
The same trait using different methods should be highly correlated, whereas different traits using the same method should not be correlated.
What is method variance?
Source of systematic error that arises from the use of a particular method e.g. Halo effect.
What is statistical conclusion validity?
Being able to affirm covariation between expected cause and effect. Causation implies covariation.
What are the problems with demonstrating covariation?
Insufficient power Satisfaction of test assumptions Controlled T1 error rate Treatment variable reliably implemented Random irrelevancies in experimental setting
What is internal validity?
Covariation is causal, no 3rd variable problems or reversal of direction.
Random assignment. Cause precedes effect.
What is construct validity of putative causes and effects?
Goal to show cause and effect relationship is about the constructs, not just particular to the measures.
Have multiple methods of performing the manipulation- if get same DV result in all then can be more confident in manipulation and theory.
What is external validity?
Demonstrated cause and effect relationship in lab generalizes to population, settings, and times.
What is the main problem validity deals with?
Trying to close the gap between what you measure and what you think it means.
