Evaluating Predictive Validity: A Rejoinder to Braden et al. (1995)

Abstract

Braden et al. (1995) argued that percentile rank scores from the subtests of SAT-d yield accurate predictive validity coefficients for the Performance scale of WISC. In doing so they have reiterated their rationale and think my investigation was flawed in some respects. I have examined their misconceptions about my study and conclude that there is nothing flawed. I have re-evaluated their rationale and find it unacceptable. I propose the study of age and other factors as moderator variables in validity inquiry with the goal of validity generalization.