What might a researcher use to establish criterion-related validity?

Criterion-related validity is a measure used to determine how well one variable predicts an outcome based on a related variable. The goal is to assess whether the test scores obtained from a measure align with the scores from an established criterion, which is often a direct assessment of the construct the test aims to measure.

Using statistical analysis of past performance is essential in establishing criterion-related validity, as it involves correlating the test scores from the newly developed measure with scores from a criterion that is already accepted as a standard. For example, if a new IQ test is developed, the researcher would compare scores from this test with scores from an established IQ test to see how well they predict each other. A strong correlation would indicate that the new test has criterion-related validity, as it successfully predicts performance on the established criterion.

In contrast, other options do not directly establish criterion-related validity. Pre-existing theoretical models are useful for framing hypotheses but do not offer direct evidence of predictive validity. Reliability coefficients assess the consistency of measurements but do not verify that the measure predicts future outcomes. Expert validation can provide insights into the content and construct validity but does not produce empirical evidence related to predicting actual performance from test scores.