Validity

In statistics and research, validity refers to how well a test or measurement actually measures what it claims to measure. A valid measure captures the intended concept accurately, not just consistently. Validity is essential for drawing meaningful conclusions from data.

In short: Does the instrument truly measure the concept it’s supposed to measure?

While reliability is about consistency, validity is about accuracy. A measurement can be highly reliable but still invalid—for example, if it consistently measures the wrong thing.

There are several types of validity, each assessing a different aspect of how well a measurement reflects the intended concept:

• Face: Does the measure appear, on the surface, to assess what it claims to?
• Content: Does the measure fully cover all aspects of the concept being studied?
• Construct: Does the measure relate to other variables as theoretically expected?
• Criterion: Does the measure correlate with an outcome or standard (criterion) it should predict?

Researchers often gather evidence for multiple types of validity to support the overall usefulness of a measurement tool.

For example, a test designed to assess mathematical reasoning should include a variety of math problems and avoid items that depend on reading comprehension. If the test focuses only on arithmetic and neglects geometry and algebra, it might lack content validity. If scores on the test predict performance in advanced math courses, it may have criterion validity. Together, these pieces of evidence help demonstrate that the test is truly measuring mathematical reasoning.