WHAT MAKES A GOOD REVIEWER AND
WHAT MAKES A GOOD REVIEW?
Susan van Rooyen, Fiona Godlee, Stephen Evans, Richard Smith, and Nick Black
BMJ, BMA House, Tavistock Sq, London WC1H 9JR, UK; London School of Hygiene & Tropical Medicine, London, UK
Objective: To examine the characteristics of reviewers in relation to quality of reviews.
Design: As part of a randomized controlled trial of blinded and "unmasked" peer review, editors, and authors were asked
to assess the quality of reviews. Demographic data were collected using a postal questionnaire, and their contribution to the
quality of reviews was assessed using multiple regression.
Results: Data were available on evaluations of 286 pairs of reviews. Increasing age was associated with lower scores (P<.0005).
Scores fell substantially (1 SD) from age 30 to 50 but did not change thereafter. Training in epidemiology or statistics was associated
with a rise in score of 0.25 SD (P=.014). North American reviewers had significantly higher scores (0.4 SD, P=.013) and were more likely (87.5%)
than all other reviewers (54.6%) to have had training in epidemiology or statistics. Each of these effects remained after adjusting for the others.
There was no detectable effect of sex, possession of a PhD or MD, academic position, number of journals reviewed for, or number of papers refereed per year.
Membership of an editorial board tended to produce lower scores. The time taken to obtain a review also showed a nonlinear effect with age, those aged <40 and
those aged >50 having shorter times. Inevitably, North Americans took longer (by 6 days on average). Data collection is continuing.
Conclusions: Older reviewers, those qualified in North America, and those with training in epidemiology or statistics produce better reviews.
Return to Session Information
