The degree to which a test actually measures what it is designed to measure.

• A. Reliability
• B. Validity
• C. Construct Validity
• D. Predictive Validity

Answer: (C)

The main idea is validity—the extent to which a test measures what it is intended to measure. The best fit here is construct validity, because it specifically asks whether the test actually taps into the theoretical construct it’s designed to assess. This includes evidence that the items align with the construct, relate appropriately to other measures of the same construct (convergent validity), and separate from measures of different constructs (discriminant validity), as well as supporting the expected factor structure. For example, a test intended to measure mathematical reasoning should relate to other math-reasoning measures and not correlate with reading tests, and its items should coherently reflect the single underlying construct.

Reliability concerns consistency of scores, not whether the test measures the right thing. Predictive validity deals with how well test scores forecast future outcomes, which is a different type of evidence. The broader term validity covers the general idea, but the description in the prompt aligns most precisely with construct validity since it focuses on measuring the intended construct.