Tuesday, July 23, 2013

Physics Departments Without A Single Female Faculty Member

Before we start, let me emphasize that I'm a man, but I've been involved in promoting women participation in science, and especially physics, for years. Anyone who has followed this blog would have read several items on this issue, and also my activities in this area to promote women in science. I definitely think that there is an under-representation of women in physics, and one of the ways to improve that is to make the field more enticing and more familiar to them, both in terms of the subject matter, and the working condition.

Now, having said that, you know what's coming next is not going to be pretty. This is a news article that reports on a recent statistical analysis/modeling done by the American Institute of Physics. It aimed to address that fact that fully 1/3 of the physics departments here in the US do not have a single female faculty member. But one shouldn't stop there, because the statistics included schools with small number of faculty members (often less than 10), and these tend to be the schools that do not have any female faculty members.

The AIP ran a simulation that takes into account the number of available female faculty members and the available positions, and came out with the conclusion that the lack of female faculty members in these departments is consistent with the statistical distribution, and not due to any inherent bias.

A new report from the American Institute of Physics -- based on simulation analysis -- concludes that the large number of departments without a single woman is to be expected and is not the result of discrimination. Some experts on women and science, however, disagree.

The institute's report says that there are two factors that explain the distribution of women among departments: the size of departments and the total number of female faculty members available. There are many departments with only two or three physics faculty members, the report notes. So "it is unlikely that these departments will have a woman among the faculty because the overall representation of women among all physics faculty members is low," the report adds.
To put it simply, say that you have 100 balls. On the table, you have many compartments of various sizes, some able to contain 20 balls, while others are big enough to have only 2 balls. If you toss those 100 balls up in the air and let them land randomly into those compartments, the argument here says that naturally, the smaller compartments will have a higher probability to end up with having NO balls.

Whether one buys into the parameters set up for the simulation is another matter. But taken at face value, I don't see anything wrong with this. It is certainly a first attempt at trying to figure out if the lack of any female faculty members in these small departments are due to some inherent bias, or simply out of statistics. It is a scientifically valid methodology to START and investigate an issue. Now the next logical step is to re-examine if the parameters used are valid, or accurate. Maybe some of the assumptions used are debatable, etc., and thus, the simulation should be tweaked.

What annoys me is the response being given to this study. I certainly expect disagreement with the conclusion, but the counter-argument that has been given is purely speculative!

Janet Bandows Koster, executive director and CEO of the Association for Women in Science, said via e-mail that the report "a disappointment."

She urged physicists to study the concept of "implicit bias," which she said might have something to do with the pool of women in the discipline. "We know that most people are reluctant to accept that they are biased, and scientists in particular pride themselves on their impartiality. Yet scientists are humans raised in societies, and thus are subject to collective messages that suggest men are suited to science because they are independent and analytical whereas women are better suited to care-giving and cooperative enterprises."

It's too easy, she said, to focus only on the relatively small number of women in the field. "Inferring there is no hiring bias because the 'n' is so small for female faculty is essentially like granting a papal indulgence to physics departments across the country," she said.
I'm sorry, but that is stupid! You are countering a statistical analysis with nothing more than a speculative fishing expedition! That's like saying you don't agree with Special Relativity because you don't like the look of the equation!

And no one here is saying that there is no hiring bias. A statistical analysis such as this can't come to such conclusion. What it does say is that the lack of female faculty members in 1/3 of the physics depts. cannot be attributed to gender bias as the main factor, because statistical analysis alone can account for that observation! As scientists, we need to know what statistics say, and what they don't!

The PROPER way to counter something like this is to look at the validity of the parameter used, to see if the model is accurate, and to show where it might have missed something, NOT to simply insinuate that there are biases. For someone who is supposed to represent an association of women in SCIENCE, she sure used a lot of hand-waving, unsupported argument to counter a scientifically-derived conclusion.


1 comment:

Tom said...

It seems to me the counter-argument was based on a different presumption, about there being more women in physics, rather than the what the study's point was, which was to use the number of women physics faculty as a starting point. Also, it should be noted that this faculty fraction is not far off from the fraction that earned PhD's a few years earlier. That would seem to indicate that the bigger issue is finding ways of getting more women into (and through) grad school.

Not to be snarky, but as Ms. Koster's argument seems unscientific, that may simply be due to the fact that she's not a scientist. (She has an MBA among other non-science accomplishments)