What is predictive validity?

What is predictive validity?

• A. The extent to which test scores predict future external criteria.
• B. The reliability of the test across time.
• C. The agreement between raters.
• D. The degree to which test scores correlate with another measure taken simultaneously.

Answer: (A)

Predictive validity is the extent to which a test score can forecast a future outcome. It’s a form of criterion-related validity where the criterion is measured after the test to see how well the scores predict it. Think of a college entrance exam predicting college GPA years later, or a job screening test forecasting actual job performance after a probation period. To assess predictive validity, you administer the test now, wait for the future criterion to be measured, and then look at the relationship between the test scores and that future outcome—usually by correlating them. The other concepts described in the distractors are about reliability or different kinds of validity. Reliability across time refers to stability of scores over repeated testing, not about predicting a future outcome. Agreement between raters concerns how consistently different scorers judge the same response. A measure taken at the same time with another criterion describes concurrent validity, not predictive validity. So the statement that best captures predictive validity is the one about predicting future external criteria.

Predictive validity is the extent to which a test score can forecast a future outcome. It’s a form of criterion-related validity where the criterion is measured after the test to see how well the scores predict it. Think of a college entrance exam predicting college GPA years later, or a job screening test forecasting actual job performance after a probation period. To assess predictive validity, you administer the test now, wait for the future criterion to be measured, and then look at the relationship between the test scores and that future outcome—usually by correlating them.

The other concepts described in the distractors are about reliability or different kinds of validity. Reliability across time refers to stability of scores over repeated testing, not about predicting a future outcome. Agreement between raters concerns how consistently different scorers judge the same response. A measure taken at the same time with another criterion describes concurrent validity, not predictive validity. So the statement that best captures predictive validity is the one about predicting future external criteria.