Peer review

Written evidence submitted by Sir John Ball (PR 24)

1. I am very concerned about the increasing prevalence of citation metrics (impact factors, h-indices and the like) as substitutes for peer review. Citations are statistics, and have some value if intelligently interpreted, but the use of such metrics as a substitute for peer review is dangerous. A good analysis of why this is can be found in the joint IMU/ICIAM/IMS report Citation Statistics (http://www.mathunion.org/fileadmin/IMU/Report/CitationStatistics.pdf ). What relevance such statistics have diminishes with the size of the sample, but we see metrics such as the h-factor being systematically used on samples of one in university promotions exercises, and letters from referees devoting more and more space to discussion of metrics rather than of the actual research done by a candidate. Among worrying trends is the increase of impact factor manipulation by journal editors through systematic efforts to increase references to papers in their journal via pressure on editorial boards, the writing of review articles etc (see http://www.ams.org/notices/201103/rtx110300434p.pdf ). 

2. Scientific journals perform important roles through quality control, improving content and presentation, dissemination and archiving. A very recent analysis by the International Mathematical Union (http://www.mathunion.org/fileadmin/CEIC/bestpractice/bpfinal.pdf) describes this in more detail and draws attention to various issues related to peer review, for example the lack of transparency in the practice of having parts of referee reports which are not sent to authors, and the ethical difficulties in ensuring there is no conflict of interest as regards peer review when the author pays for publication.

3. (Declaration of interest.) I was one of the authors of the report mentioned in paragraph 2 and I am the editor of various academic journals.

Sir John Ball

Mathematical Institute

Oxford

3 March 2011