Etan-main text

Science policies in the European Union
grant a male applicant a post-doctoral position than a female one.  The
applications for awards for postdoctoral positions in 1995 (52 women and
62 men) were investigated.  The ‘scientific competence’ of the applicants had
been judged on a scale from 0 to 4 by the MRC evaluators.  Women
obtained lower scores on average for scientific competence compared to
their male colleagues.  Using multivariate analysis, the competence score
given to an applicant was related to a number of characteristics of that
applicant, including scientific productivity (number and impact factor of
papers, citations in scientific press, etc.), gender and research field.  Three
factors were found to be independent determinants of high scores for
‘scientific competence’:
•
the applicant’s scientific productivity;
•
gender (male applicants received higher competence scores than female
applicants with equal productivity), and;
•
affiliation with one of the peer review committee members (applicants
with such ties, who had been supervised by one of the evaluators for
example, obtained better scores than other applicants exhibiting equal
productivity).
The extra competence points allotted male applicants because of their
gender corresponded to 20 extra scientific publications in excellent
specialist journals.  Thus, compared with the average male applicant, a female
scientist had to be 2.6 times more productive if both were to be perceived
as equally competent (Figure 4.1)
The above study was published in Nature in 1997 under the title ‘Nepotism
and Sexism in Peer Review’ (Wennerås and Wold, 1997).  The paper caused
extensive reaction both in the scientific and the popular press.  The board of
directors of the Swedish MRC was replaced, in part because of the Wold-
Wennerås study.  The proportion of female evaluators was increased and
strict guidelines and monitoring procedures were introduced to reduce
injustice in the evaluation procedure.  The end result was increased fairness,
and hence, quality in the choice of recipients of grants and positions.
Another key factor in the MRC’s improvement of the peer review system
was increased transparency of the evaluation process.  Thus, all applicants
now routinely receive their evaluation scores and the identity of the peer
reviewers is known.
Scrutiny of peer review in the Netherlands
The publication of the Wennerås and Wold (1997) article inspired a similar
study in the Netherlands, sponsored by the Dutch Ministry of Arts and
Sciences and conducted by the Dutch Association of Women's Studies.  The
researchers analysed success rates of male and female applicants to the main
research granting agencies, the Netherlands Organisation for Scientific
Research (NWO) and the Dutch Royal Academy of Arts and Sciences
(KNAW).  They concluded: ‘gender does play an independent role in the
evaluation of women, regardless of their academic merit’ (our emphasis).
The evaluation of men, however, was found to be in tandem with their
academic status (Brouns, 1999).
The researchers had addressed the question: ‘Is it possible to identify any
gender differences in the evaluation systems and the awarding of grants by
the NWO or the KNAW?’ They conducted a survey that focused on the
post-doctoral fellowship programme of the KNAW and two programmes
‘... the credibility of the
academic system will be
undermined in the eyes of the
public if it does not allow a
scientific evaluation of its own
scientific evaluation system.’
Wennerås and Wold ,1997, p 341
2.9
2.8
2.7
2.6
2.5
2.4
2.3
2.2
2.1
2.0
0-
19
20-
39
40-
59
60-
99
>99
Total impact
Competence scor
e
Men
Women
Figure 4.1: Mean competence
scores given to male and
female applicants by the
Swedish MRC
Note: The mean comptence scores
are plotted as a function of scientific
productivity, measured as total
impact. One impact point equals one
paper published in a Journal with an
impact factor of 1.
Reprinted with permission from
Nature, vol 387, pp 341-3 (1997)
34

---

for individual research grants within the NWO: the Talent-stipendium and
the Pioneer-programme.  They found that the percentage of women who
applied was, in all the programmes, by and large, similar to the percentage of
grants awarded to women.  However, a detailed comparison of some of the
applications was conducted to test the hypothesis that women applicants
might have a better track record than men, since they had already survived
the gendered selection processes in academia.  An in-depth analysis of 138
Talent applications and a selection of files in the years 1993 and 1994 in two
disciplines were scrutinised: the exact sciences (E) and the biological, ocean
and earth sciences (BOE).  The disciplines were selected for being
remarkable: in the exact sciences, women’s success rate was significantly
better than that of men (100%), in the biological, ocean and earth sciences,
it was significantly worse (0%).
Table 4.1:  Analysis of applications to the Dutch research bodies by gender,
1993 and 1994
Programmes
Male
Male
Female
Female
applicants
awards
applicants
awards
Talent total
188
80
43%
35
13
37%
E sciences
81
34
41%
8
8 100%
BOE sciences
36
10
28%
9
0
0%
The analysis, based on multiple regression, revealed interesting tendencies in
the relationships between characteristics of applicants (productivity, age and
promotion speed), evaluation by external advisers, and decisions of the
NWO (however, the numbers prevent statistically significant conclusions).
On the level of the whole population, the figures are reassuring: the decision
of NWO is in concordance with the external evaluations.  However, there is
a problem.  Statistically the evaluations of the external advisers are unrelated
to the characteristics of the applicants and hence the basis for the evaluation
is unclear.
It should be noted that of the 270 external advisers (peers) who were
involved in the evaluation of the applications, only 4 were women.
When the NWO decisions are analysed with respect to gender, it is
remarkable that for men, the decisions correlate with the productivity, age
and promotion speed of the applicants.  However, with respect to women,
this was not the case.  Women applicants in the E-sciences were generally
evaluated as better than men, but seemed to receive a bonus.  Women in the
BOE sciences were generally evaluated to be as good as men, but the
decisions were negative.  In the evaluation of the external advisers of the
BOE applications, there appears to be a gender bias as well.  Women and
men with equal track records receive different evaluations.  Men are much
more often marked as excellent than women (Brouns, 1999).
Peer review in the UK
The MRC in the UK also analysed application and award rates for their
fellowship schemes by gender for the year’s 1993/4 to 1996/7. For the
clinical fellowships, women comprised 31.9% of applicants but received 33%
of the awards. For the clinician scientist fellowships, women were 24.8% of
the applicants but obtained 32.5% of the awards. For the non-clinical
research fellowship scheme, women were 50.8% of the applicants and got
Fairness and funding/modernising peer review
Correcting discrepancies
In Sweden, the Medical
Research Council (MRC)
retrospectively investigated
their grant-giving policy with
respect to gender and seniority
of the researcher. Among the
scientists given equal
competence scores by the
MRC peer reviewers, male
researchers obtained larger
grants than female scientists.
Older scientists obtained larger
grants than younger scientists.
The Swedish MRC has
corrected these discrepancies
and the size of the grants is
now only based on competence
scores (Abbot, 1997).
35

---