# Biased journals

Just read a fascinating little article called “Topical Bias in Generalist Mathematical Journals” by Joseph Grcar published in the Notices of the AMS.

Its premise is very simple – examine the bias (carefully defined as referring *only* to statistical deviation from the expected rather than its common more emotive meaning!) of the generalist journals published by the AMS. For each subject area (as determined by 2-digit MSC code), the author compares the proportion of articles in that subject area published by, say, “*Proceedings of the AMS*“, with the proportion of articles in that subject area across all mathematical output (as indexed by Zentralblatt).

The results show a *vast* variation between the subject areas. For example, from 2000-2009, 6.5% of the papers in the *Proceedings* concerned Group Theory (MSC 20) compared to 2.6% across all mathematics, thus indicating a *positive bias* in favour of Group Theory. Although this is the example Grcar quotes in the text, his tables reveal a huge range, with the proportion of papers on Commutative Rings (MSC 13) appearing in the Proceedings about 6 times higher than the overall proportion of papers on Commutative Rings and many (particular Applied Maths) areas occurring in the *Proceedings* at a much lower rate than their overall proportion.

Of course this is only a small study, and only for a couple of journals and there are all sorts of reasons why bias in this statistical sense may occur that do not involve bias in the *malicious* sense, but I’m sure that lots of people will be reading this with great interest, especially those who feel (rightly or wrongly) that their good work is hindered by working in an “unsexy” area of mathematics! Whatever eventuates from this, Grcar has certainly managed to write a great discussion starter for Mathematical coffee rooms around the world.

Combinatorics (MSC 05) by the way, seems to be only mildly underrepresented in the *Proceedings*, so while we’re no Algebraic Topologists (MSC 55, overrepresented by factor of 4) at least we’re not working on Classical Thermodynamics (MSC 80, underrepresented by a factor of 16+)!