OBJECTIVE: To analyse whether physician assessment of scientific quality is biased by gender. DESIGN: Two fictive research abstracts on back pain treatment were constructed, one with a quantitative and one with a qualitative design. Authorship was assigned to either a woman or a man. SUBJECTS: 1637 randomly selected Swedish physicians were asked to judge the scientific quality of the two designs in a structured assessment form. MAIN OUTCOME MEASURES: The assessments of 1364 abstracts (286 female and 394 male assessors) were analysed by chi-square test and logistic regression. RESULTS: The quantitative design was judged the same, regardless of the gender of the author or assessor. The qualitative design, however, was ranked as more accurate, trustworthy, relevant and interesting with a female author. Women assessors upgraded female authors more than male authors, while male assessors reflected no gender differences. Assessor speciality interacted with judgement; physicians in primary care appreciated the qualitative abstract more than hospital physicians did (OR 2.78; 95% CI 1.97-3.92). CONCLUSION: Gender seems to affect scientific evaluations. The results are worth considering in situations where research is judged and interpreted, in medical tutoring, research guidance, peer reviewing and certainly in forming evaluation committees for research funding.
OBJECTIVE: To analyse whether physician assessment of scientific quality is biased by gender. DESIGN: Two fictive research abstracts on back pain treatment were constructed, one with a quantitative and one with a qualitative design. Authorship was assigned to either a woman or a man. SUBJECTS: 1637 randomly selected Swedish physicians were asked to judge the scientific quality of the two designs in a structured assessment form. MAIN OUTCOME MEASURES: The assessments of 1364 abstracts (286 female and 394 male assessors) were analysed by chi-square test and logistic regression. RESULTS: The quantitative design was judged the same, regardless of the gender of the author or assessor. The qualitative design, however, was ranked as more accurate, trustworthy, relevant and interesting with a female author. Women assessors upgraded female authors more than male authors, while male assessors reflected no gender differences. Assessor speciality interacted with judgement; physicians in primary care appreciated the qualitative abstract more than hospital physicians did (OR 2.78; 95% CI 1.97-3.92). CONCLUSION: Gender seems to affect scientific evaluations. The results are worth considering in situations where research is judged and interpreted, in medical tutoring, research guidance, peer reviewing and certainly in forming evaluation committees for research funding.