Once again, the folks at hangingtogether are my source for new research — in Efficiency and scholarly information practices, they describe a new study published in Journal of the American Society for Information Science and Technology : JASIST that purports to refute an earlier study published in Science magazine that shows that as journal content becomes available online, citations get narrower.
(Full citation: See: Larivière, Vincent, Yves Gingras, and Eric Archambault. “The Decline in the Concentration of Citations, 1900-2007.” Journal of the American Society for Information Science and Technology : JASIST. 60. 4 (2009): 858-862. Stanford-only link).
The original article “Electronic Publishing and the Narrowing of Science and Scholarship”, by friend and colleague James Evans, attracted quite a bit of attention (I wrote a bit about it myself), so I am not surprised that someone has already tried to replicate (or refute) his findings.
In “The Decline in the Concentration of Citations, 1900-2007.” the authors claim that contrary to Evans’ findings, citations are becoming increasingly dispersed over time. The major problem here is that Evans has done a fairly sophisticated analysis (including negative binomial models) that controls for time. This allows him to isolate the estimated effect of journal online availability on trends and patterns of citations. In contrast, Lariviere, Gingras, and Arhambault readily admit that their models “do not take into account ‘the online availability’ variable.”
In other words, the study that claims to refute Evans’ work actually fails to account for the key independent variable, and instead shows simple trends over time.