Construct validity

Construct validity is evidence that an assessment actually measures the abstract concept ("construct") it claims to measure. It's typically established through patterns of correlation with related and unrelated measures: a depression scale should correlate strongly with other depression measures (convergent validity) and less strongly with measures of unrelated constructs like extraversion (discriminant validity).

Construct validity is the most rigorous form of validity and is built up gradually across many studies, since the construct itself isn't directly observable.