May 23, 2005

A Sociological Exercise

Correspondence with a number of people who, for obvious reasons, I won't name, suggests that part of the reason physicists so much bad biology, sociology, etc. shows up in physics journals, is that it is "too easy" for physicists to slap together some shoddy simulations and think they're answering important questions, especially if they know very little about the field they're invading. My correspondents further suggest that the people who do this are likely to be the ones who couldn't succeed as ordinary physicists. This case is to be distinguished, they suggest, from physicists who become interested in other fields, take the trouble to learn about them, and, most likely, end up publishing in their fora as well as, or even instead of, physics journals.

Let me suggest, then, the following exercise, which should be fairly straightforward, if quite tedious, for anyone with access to the right databases and a copy of R. Take, say, Physica A over the last few years, and identify all the authors of papers on non-physics subjects. (I suggest counting quantum computation as physics, at least for the purposes of this study, for reasons I hope to explain later.) Divide these authors into the conquistadors, who only publish in physics journals, and the assimilated, who also publish in relevant non-physics journals. For each group, compare their publication record in conventional physics, as measured by say citations per year per paper, with that of straight physicists, who only publish on conventionally-physical topics. The hypothesis is that the conquistadors will have been less successful, within physics, than either the straight physicists or the assimilated. One should probably control for time since receiving the doctorate, and possibly for for the school they got their doctorate from as well.

If you do this study, let me know the results, OK? I'm curious.

Physics; Learned Folly; Modest Proposals

Posted at May 23, 2005 18:00 | permanent link

Three-Toed Sloth