Performance tests in human psychopharmacology (2): Content validity, criterion validity, and face validity

Abstract

AbstractThe four traditional aspects of validity comprise: criterion, content, face and construct. Criterion validity is assessed by calculating the degree of correlation between test score and external criterion (e.g. driving ability of flying skill). While this form of evidence is rare, it has been demonstrated in a few psychopharmacology studies, especially with psychomotor tasks (e.g. target tracking). Content validity is an index of the appropriateness of test selection. Many test batteries cover a wide range of psychological functions, and thus display a degree of content validity. However, the functions being assessed often differ, reflecting the absence of agreement over what ‘human performance’ should comprise. Most test batteries also contain a preponderance of simple tasks, and content validity could be improved through the inclusion of ‘higher cognitive’ tasks. The difficulties of assessing performance using a single test are also critically examined. Face validity is not a true index of validity, but a reflection of whether the test looks right. It can often be misleading, although it does have positive functions (e.g. for subject motivation). Sensitivity to psychoactive drug effect is also a prerequisite for tests in this area, although it is not in itself an index of validity. The construct validity of human performance tests, together with an overview of test meaning and interpretation, are examined in the final paper of this series.