Is there gender bias in research fellowships awarded by the NHMRC?

Applications by researchers for appointment, reappointment or promotion are highly competitive. Criteria used to evaluate applications include the applicant's independence and track record in research, originality of the research, national and international recognition, publications and broader contribution to the applicant's area of research. These criteria are not weighted: rather, the overall merit of each case is assessed from diverse sources of evidence such as curriculum vitae, reports from referees nominated by the applicant, reports from external assessors nominated by the Research Fellowships Committee, consideration of the regional grants interviewing committee score (which indicates the scientific quality of the project or program to which the fellowship application is tied) and interview.

A Swedish study showed that reviewers' scores of postdoctoral fellowship applications to the Swedish Medical Research Council were strongly influenced by the gender of the applicant.² This prompted the Research Committee of the NHMRC to ask the Research Fellowships Committee to conduct its own review.

Analyses were not conducted for reappointments as we could not be confident that these would not include repeated applications from individuals who had failed in an initial application or an application for promotion within the same period.

Over the study period, 81 applications for initial appointment at Research Fellow level were received from men compared with 45 from women (ratio 1.8 : 1). Similarly, 121 applications for initial appointment to a Senior Research Fellowship or higher were received from men compared with only 38 from women (ratio 3.2 : 1); 30 applications for initial appointment as Principal Research Fellow or Senior Principal Research Fellow were received from men and only four from women (ratio 7.5 : 1); a total of five applications for initial appointment as Senior Principal Research Fellow were received from men compared with only one from a woman (ratio 5 : 1).

We combined applications for appointment as Research Fellow or Senior Research Fellow and compared these by sex with those for Principal Research Fellow or Senior Principal Research Fellow. Of the 202 applications received from men, 30 (15%) were for Principal Research Fellowships or Senior Principal Research Fellowships. For women, only 4 (5%) of 83 applications were for initial appointment as Principal Research Fellow or Senior Principal Research Fellow. Having applied for an initial appointment, men were significantly more likely than women to apply for a senior appointment (² = 5.6, df = 1, P = 0.02).

For the study period, 26 applications (13%) from men for initial appointment (irrespective of level) were successful and 176 (87%) were not. In contrast, 14 (17%) applications from women for initial appointment (irrespective of level) were successful compared with 69 (83%) unsuccessful. The difference between men and women is not significant (² = 0.8, df = 1, P = 0.4).  

For the study period, 34 (48%) applications from men for promotion (irrespective of level) were successful and 37 (52%) were not, out of the total of 71. In contrast, 14 (58%) applications from women for promotion (irrespective of level) were successful compared with 10 (42%) unsuccessful, out of the total of 24. Again, there was no significant gender effect (² = 0.8, df = 1, P = 0.4).

Our study also shows that, among new applicants, men are more likely than women to apply for fellowships at the higher levels. The data might also suggest that male research fellows are more likely to seek promotion to the higher levels than female research fellows. However, these analyses were not adjusted for age or years of postdoctoral experience, so our data on potential gender bias in promotion are very limited. To obtain better data, it would be necessary to select a cohort of research fellows appointed in one year and track their progress, testing statistically whether gender is associated with further applications for promotion or reappointment.

The Research Fellowships Committee itself has five male and four female members.¹ In keeping with increasing community and professional interest in the accountability of the NHMRC,^1,4-6 we place our findings in the public domain to generate discussion. We conclude that the influence of gender bias, if present at all in the research fellowships scheme, is indirect and acts before the process of evaluation of a specific application. Women may be less likely to apply and, once appointed, perhaps more cautious in their applications for promotion. However, having applied and specified a particular level, the gender of applicants does not influence the outcome of their application.

We acknowledge the limitations of the available data. Access to applications to calculate publication outputs and acquisition of competitive grants as indicators of research proficiency (as in the Scandinavian study)² would have required the written consent of the applicants. Further debate and resources to support such a study are recommended, as is research to examine any differential success rates in project or program grant applications to the NHMRC by male and female investigators.

1. National Health and Medical Research Council. 1997 Annual Report. Canberra: NHMRC, 1998. (Commonwealth of Australia Catalogue No. 9804863.)
2. Wenneras C, Wold A. Nepotism and sexism in peer review. Nature 1997; 387: 341-343.
3. Department of Employment, Education, Training and Youth Affairs. Selected Higher Education Student Statistics, 1997. <http//www.deetya.gov.au/divisions/hed/ highered/statpubs.htm>. Accessed 18 September 1998.
4. Anderson W. Funding Australia's health and medical research. Med J Aust 1997; 167: 608-609.
5. Ward J, Slaytor E. Enhancing NHMRC investment in public health research. Aust N Z J Public Health 1998; 22: 189-190.
6. Larkins R, Anderson P. Australian medical research: more resources and the right balance. Med J Aust 1998; 168: 535-536.