Refereeing Reconsidered: An Examination of Unwitting Bias in Scientific Evaluation

Abstract

This paper argues that while the majority of referees nay attempt to give unbiased evaluations, they must inevitably bring their own particular sets of scientific predispositions to bear upon the recommendations they offer to journal editors. Qualitative and quantitative evidence is presented which demonstrates that resulting biases are systematically observable in physical as well as social and behavioural science.

It is concluded that editors must be sensitive to the intellectual affiliations and preferences of potential referees, and bear these in mind when selecting evaluators and interpreting their reports.