Relationship of Department Size and Publication Productivity 3 . Relationship of Publication Productivity and Department Size to Doctoral Program

This study extends previous work relating contemporary approaches to the comparative evaluation of doctoral programs, focusing on correspondences between the 2004 NCA study, whose assessment strategy was based solely on perceptual measures, and the new ComVista system, whose assessment is based on behavioral measures – departmental publication patterns and publication productivity. The ComVista publication data are coded for topical content and cluster analysis of these codes revealed 17 clusters of intellectual activity in the field. A second cluster analysis grouped doctoral programs on the basis of the frequency and distribution of their publication output in the 17 areas, revealing nine types of doctoral programs differentiated in terms of the intellectual directions of faculty publications. Additional analyses examined the ability of the 17 ComVista topical content clusters to predict the NCA ratings for perceived quality of doctoral faculty finding substantial prediction. Frequency of publication in the content clusters accounted for as much as 70% of the variance in the NCA ratings, suggesting that these data, which are often regarded as indicators of the relative prestige of doctoral departments, are principally based on peer perceptions of faculty publication activity. Other results suggest that doctoral programs that are more specialized in the direction of their publication activities tend to be perceived more favorably, and that the number of faculty in a program is strongly related to positive perceptions of faculty quality, provided the number of nonpublishing faculty in the