Assessment centers are extraordinarily popular among professionals and practitioners involved in the selection of managerial personnel. This popularity is reflected in a large number of published articles (e.g., Finkle, 1976;Huck, 1976;Klimoski & Brickner, 1987), the publication of standards and ethical considerations involved in assessment center operations (Task Force on Development of Assessment Center Standards, 1977), and the unusual support of the federal courts (Byham, 1979). This enthusiasm is supported by a great deal of validation research. Over 50 studies were reviewed by Huck (1976), all of which reported positive findings concerning the relationship between assessment center ratings and subsequent job performance. Klimoski and Strickland (1977) found a median validity of .40 over 90 studies.
More recently Gaugler, Rosenthal, Thornton, and Bentson (1987) reported a meta-analysis of assessment center validity. These authors calculated average validities for four different types of criteria. Results indicated the strongest evidence of validity occurred when assessment center performance was correlated with ratings of managerial potential (.45). Somewhat lower validities were found when the criteria used were ratings of job performance (.31), rating of the manager's job performance on the dimensions used in the assessment center (.25), the performance of a manager in a training program (.31), or career advancement (.32). Overall, Gaugler and associates (1987) reported the average validity of assessment centers to be .37.
Research evidence regarding the application of assessment centers in the educational context is much less common. In fact, a recent review of the literature on principal selection indicated only a handful of studies that provided any empirical data evaluating the utility of procedures used to select administrators (Schechtman & Schmitt, 1987). Ehinger and Guier (1985) used a concurrent validation strategy to evaluate the Management Development Program at the University of Tulsa. These authors developed a composite assessment center score representing performance on the in-basket and the leaderless group discussion and found a moderately strong relationship with superintendents' ratings of job performance (r= .41, n = 103).