Key points are not available for this paper at this time.
What would a strong program of construct validation look like for the concept of test anxiety? What are the components of strong validation programs? In particular, how does structural equation modeling fit into such a program?
Jeri Benson (1998) studied this question.