Citation ‘Bias’ in a Small World

Image: Breakdown of citations on the KamLAND paper

The Nature paper I was given to analyse, ‘Experimental investigation of geologically produced antineutrinos with KamLAND‘, turned out to have one of the highest rates of citation bias of any of the papers we examined.

 As Debora Miranda puts it, citation bias refers to  ‘a potential influence of choice’. Bias A means one of the authors of the original article was the same as in the citing paper. Bias B means he or she came from the same institution as an author in the citing paper. And Bias C refers to any other known link between the authors.

Hopefully we have made it clear that, although a high level of bias may imply an element of self-promotion or cliquishness, we can’t draw such inferences lightly. This paper is a case in point. Despite the fact that 33% of the citing papers showed some ‘bias’, there are legitimate reasons why this should be the case.

Detecting Geoneutrinos

The paper I was given was, on first read, almost indecipherable to a non-specialist. Published in Nature in July 28 2005, it consisted of an impenetrable mass of equations and geophysics jargon, a sample sentence being:

For typical geoneutrino energies, the approximation P(E,|L|) = 1 – 0.5 sin²2 ₁₂ only affects the accuracy of the integral in equation (1) at 1% owing to the distributed _eproduction points.

This does not make for a light and easy bedtime read (although my cynical first impression was that it might send me to sleep). Moreover, with a daunting 87 authors to sift through, citation checking looked to be quite a task.