Thanks again for your comments.

You are right to identify that the question we ask and how we answer is it is the crux of the matter. If productivity is the main benchmark of research excellence, as is the case at present, then of course funding/promotions are going to be judged on this criterion, regardless of gender of the applicant. I won't reiterate the reasons why this may be unfair to women. We undoubtedly want to reward the 'best' scientists, but my question is - what is the best measure of 'best'? We don't think it strange in science to consider productivity as the key indicator of performance. Yet in the arts it would be considered outrageous to award, say, government funding to a painter based on how many pictures he or she had produced in the last five years. We'd be much more concerned with the 'quality' of their work (I appreciate that this is a much more nebulous concept in the arts). Obviously there needs to be a base line of productivity - it would be foolish to fund someone who never produces anything - but the overwhelming importance of quantity of output in science (think how we are encouraged to divide our work into least publishable units) is I think problematic, particularly if some elements of the workforce face obstacles in achieving high productivity.

I agree that improving the standing of any under represented group should not be automatically done just because they are under represented. But it is important to understand WHY a group is under represented in order to assess whether something should be done.

My dictionary (Chambers' 20th Century) defines impartial as 'not favouring one more than another'. In the context of discussing the use of research productivity as a metric of research excellence, then it is appropriate since males are favoured by this measure.

S-E-L

Apologies for the delay in replying. Though my sprint coach analogy was a bit strained, I disagree with the assertion that it was disingenuous, actually because of the very same reasons that you state; i.e. what are deemed to be the important criteria for selection. My point was to call into question the heavy focus on one parameter (running speed in the analogy, quantity of publications for real) that may be neither important* nor impartial to assessing ability. Beyond the obvious point of not wanting to fund people who don't get anything achieved, why should quantity be valued so highly?

This question becomes even more pertinent if confronted by the possibility (we would argue likelihood) that a proportion of the scientific community face obstacles that prevent them from achieving high productivity (obstacles that are not related to their own ability).

It would be interesting to know exactly why productivity is correlated with quality of publications. How could one test whether it reflects overall research ability (the best scientists produce high quality science in large amounts), as you suggest, or whether people who publish lots tend to get cited more because their name is more prevalent (the ‘fast food effect’ suggested by Kelly and Jennions – ref 9 in our paper)?

If you assume the former, then by extension you are saying that those who are not as productive are in fact less able scientists. The point of our Research Status metric is to control for a variable that we do not want to measure, on the basis that it does not reflect ability and discriminates against women scientists (both points, I concede, with which some/many will disagree). Controlling in this manner (using residuals) for unintended variables is standard practice in allometry. I was going to extend the basketball analogy here, but I suspect it will start to cloud the issue, so I'll leave it at that.

* I should qualify the word important here. From the point of view of a university department deciding on a job appointment (for example) then it is difficult to criticise them for taking strongly into account the productivity of the candidate. Currently, that is the key indicator of their future funding success. Whether, though, at the funding body/research assessment level and more generally, productivity is an important measure of scientific ability, is debatable.