Psychological Testing Chapter 5
Face Validity
Accepting that it is valid, although it may not be evidence based it is not validity
Construct Irrelevant Variance
Noise. The test measures more than just the construct
Validity Coefficient
The correlation coefficient that results from relating the test being evaluated w/ the criterion measurement
Predictor
Used to predict future outcome
5 Sources of Evidence
1. Based on content 2. 3. Based on internal structure 4. Based on Response Processes 5. Based on consequences of testing
Validity
A agreement between a test score or measure and the quality it is believed to measure
Convergent Evidence
Evidence showing that you're measuring the same thing
Construct Related Evidence
Exists when a test adequately represents the universe of items it is intended to sample from
Criterion Validity
If the test predicts performance of another outcome
Subjective
Includes personal feelings, own thoughts and opinions, can be more prone to unreliability and to criterion contamination
Criterion
Is the future outcome of performance
Concurrent Method
Is used to account for variability in performance on an outcome (criterion) that is measured at the same time
Objective
May be too limited, may not have a opinion, inadequately capture the construct of interest. More of a factual perspective
Discriminate (Divergent) Evidence
Measuring unrelated construct
Things you should consider when choosing variables
Objective & Subjective
Construct Under-representation
Test measures less than all aspects of the construct
Predictive Method
Test test is used to forecast performance on an outcome (criterion) that is measured at a later time
Content Validity
Whether the items on a test actually represent the domain of the test