Information Research Weblog

Quotation from an extract:

Dr. Liolios: At MEDNET 2005, you reported some interesting data on the citation rate of OA articles, showing that OA articles are cited more often than non-OA articles.

Dr. Eysenbach: The OA study was a longitudinal bibliometric analysis of a cohort of OA and non-OA articles. While there have been other studies before that claimed that OA articles are more frequently cited, these previous studies all suffer from huge methodologic problems because they just compared crude citation counts of openly accessible articles on the Internet. You can’t just compare OA articles vs non-OA articles without adjusting for the many different confounders. To my knowledge, the study I presented is the first rigorous study that applied multiple regression techniques to adjust for the many possible confounders -- and it still found, after adjustment, that OA articles are 3 times more likely to be cited than non-OA articles in the first 10-16 months after publication. This is clear evidence of the fact that OA accelerates the speed with which new findings are taken up by peers. It ultimately speeds up the pace of progress and knowledge translation....

Read the extract at Peter Suber's site

My only critical comment on this quotation is that Dr. Eysenbach, like many others, assumes that there are only two alternatives: author pays, and subscription journals. There are three, the third being the subsidised journal - like Information Research. Until the subsidy model is recognised as a viable 'third way', the goal of true open access will not be achieved.

Later in the extract, Dr. Eysenbach noted that the notion that OA journals were likely to have less rigorous peer-review processes was nonsense: he might actually have said that they were likely to be higher, especially in relation to the subsidised, free journal. A print journal has pages to fill, whereas a free, subsidised, electronic-only OA journal can be highly selective, since it can publish several papers or one paper in an issue.