Which validity ensures that a measure does not relate to measures of different theoretical constructs?

Which validity ensures that a measure does not relate to measures of different theoretical constructs?

• A. Discriminant validity
• B. Convergent validity
• C. Criterion validity
• D. External validity

Answer: (A)

Discriminant validity focuses on distinctness between constructs. When you have a measure for a given psychological trait, you expect it to relate to measures of the same or closely related constructs (demonstrated through convergent validity) but not to measures of different, unrelated constructs. If the new measure shows only low or negligible correlations with tests tapping different theoretical constructs, that supports discriminant validity—the measure is capturing something truly distinct rather than overlapping with other concepts. For example, a scale intended to assess anxiety should not strongly correlate with a scale measuring physical fitness; the low relationship supports that the anxiety measure is not just picking up unrelated attributes. In contrast, convergent validity would be shown by high correlations with other measures of the same construct, criterion validity by relationships with external criteria (like predicting an outcome), and external validity pertains to generalizability of findings beyond the study.

Discriminant validity focuses on distinctness between constructs. When you have a measure for a given psychological trait, you expect it to relate to measures of the same or closely related constructs (demonstrated through convergent validity) but not to measures of different, unrelated constructs. If the new measure shows only low or negligible correlations with tests tapping different theoretical constructs, that supports discriminant validity—the measure is capturing something truly distinct rather than overlapping with other concepts. For example, a scale intended to assess anxiety should not strongly correlate with a scale measuring physical fitness; the low relationship supports that the anxiety measure is not just picking up unrelated attributes.

In contrast, convergent validity would be shown by high correlations with other measures of the same construct, criterion validity by relationships with external criteria (like predicting an outcome), and external validity pertains to generalizability of findings beyond the study.