What does validity measure?

What does validity measure?

• A. Does the test test what it is supposed to test?
• B. Is the test reliable?
• C. Does the test have good item discrimination?
• D. Does the test have a fair scoring rubric?

Answer: (A)

Validity measures whether the test measures what it’s supposed to measure and whether the resulting scores support appropriate interpretations. In practice, this means evidence that the items tap the intended construct (for example, problem-solving skills when that is what the test aims to assess) and that inferences drawn from the scores are meaningful and appropriate. Reliability, by contrast, is about consistency of results across items or administrations, and a test can be reliable without being valid if it doesn’t target the right construct. Item discrimination shows how well individual items differentiate between higher- and lower-ability students and helps with test quality, but it isn’t the measure of validity itself. A fair scoring rubric ensures consistent scoring and fairness, yet that alone doesn’t establish validity. So validity is specifically about whether the test truly measures what it is intended to measure.

Validity measures whether the test measures what it’s supposed to measure and whether the resulting scores support appropriate interpretations. In practice, this means evidence that the items tap the intended construct (for example, problem-solving skills when that is what the test aims to assess) and that inferences drawn from the scores are meaningful and appropriate. Reliability, by contrast, is about consistency of results across items or administrations, and a test can be reliable without being valid if it doesn’t target the right construct. Item discrimination shows how well individual items differentiate between higher- and lower-ability students and helps with test quality, but it isn’t the measure of validity itself. A fair scoring rubric ensures consistent scoring and fairness, yet that alone doesn’t establish validity. So validity is specifically about whether the test truly measures what it is intended to measure.