Conventional wisdom holds that presidents these days are selected for their skill as fund raisers and lobbyists, not for anything so mundane as original scholarship. Sure, a Ph.D. is still a requirement at most institutions, but in an era of career administrators, a presidential bio is supposed to boast of capital campaigns not journal articles.

Actually those who long for the days of universities led by real scholars may be surprised by a new study that found a correlation between being a well respected (and published) researcher and obtaining a top presidency.

Amanda Goodall, a researcher at the business school at the University of Warwick, in Britain, examined all of the research citations of work done by presidents of 100 top universities all over the world. The research, forthcoming in The Journal of Documentation, classified all the presidents as those with high or low research expertise based on how frequently their research work is cited. (Goodall used various conversion methods to deal with the way some disciplines are cited much more frequently than others, and presidents were assigned a category on whether they were cited highly or not.)

Goodall, while acknowledging flaws in university ranking systems, used the top 100 list done by a research team at Shanghai Jiao Tong University. (Its top 10: Harvard, Stanford, Cambridge, Berkeley, MIT, CalTech, Princeton, Oxford, Columbia and Chicago.) The presidents whose academic records were studied were in office at the end of 2004.

What Goodall found -- in both American and non-American institutions -- was a significant correlation between the quality of research done by presidents and how high up on the prestige ladder they were situated. The best universities have as presidents people with the most distinguished scholarly records.

In terms of citations, those leading the top 50 universities (a group that is made up primarily of American institutions) are two and a half times more likely to be cited than presidents in the next group of 50. And a president in the top 20 (of which 17 are American universities) has almost five times the citations of a president in the fifth quintile.

In her paper, Goodall offers several possible explanations. Among them: good researchers may make good presidents, top universities see good researchers as being assets to the institution in fund raising and public relations, and research talent may be a proxy for leadership skills. Goodall's paper makes clear that there is not yet evidence for any one of those explanations (or other possibilities).

In an e-mail interview, said that "the most significant part of these findings is the fact that this is an important characteristic about leaders of research universities that has until now not been picked up on or analyzed."

While noting that managerial skills are obviously important for university presidents, Goodall said that she was pleased with the correlation she found. "It could be argued that a leader should reflect the mission statement of a university and as we can see with the top ones, it does," she said. "Being a good researcher also reflects the core business of the research university."