Abstract
In an invited address, R. M. Thorndike recently provided a critical review of several factor analytic studies that address the construct validity of the Stanford-Binet: Fourth Edition. After reviewing the studies, he characterizes a particular set of factor analytic strategies he used as representing the “right” choice. In this article I briefly review techniques for establishing construct validity in general and his arguments in particular. I conclude by recommending consideration of a confirmatory factor analysis followed by related hypotheses testing procedures (e.g., the LISREL program) for the purpose of establshing model-specific construct validity.