An analysis of the determinants of salary structures in American universities, based on data from a national survey, reveals higher salaries for those who publish than those who don't, lower salaries for those teaching than for researchers or administrators, and substantial variation in salaries in various fields. (JT)

Administrators wishing to stimulate research productivity should try to lure the most creative researchers but should also put emphasis on libraries and information retrieval services, and funds to attend professional meetings. (Editor)

The motivations and costs of collaboration in research reflect tensions between scientists' individualistic dispositions and the communal norms of science. Collaboration affects the advancement and integrity of research, and assessment and evaluation help determine the conditions under which collaboration is in the best interests of scholarship.…

From this data it would appear that teacher ratings are only partly due to individual differences in teaching abilities but also vary with posi tion within the social structure of the university. Some of the conflicting implications of the research findings for university policy are considered. (Editor/PG)