MEASUREMENT INVARIANCE IN LONGITUDINAL CLINICAL RESEARCH ASSUMING CHANGE FROM DEVELOPMENT AND INTERVENTION

Valid and reliable measures of psychological and behavioral constructs are critical to clinical research, particularly longitudinal treatmen t research, in which multiple groups are compared over time for possib le changes within and between constructs as a result of intervention o r development. Structural equation modeling (SEM) analysis is a class of statistical procedures that can be used to test multiple hypotheses about these relationships simultaneously while controlling for measur ement error. The procedures have been applied primarily to testing bet ween-construct relationships in nonexperimental studies, with relative ly little emphasis on establishing whether measures are sufficiently i nvariant across groups and across time to permit these tests. This art icle uses an empirical example of a longitudinal experimental preventi on study with two groups to illustrate the use of SEM, first, to syste matically test measurement invariance across groups at each wave of me asurement, and second, after establishing measurement invariance, to t est structural invariance longitudinally.