In this post from a few weeks ago, I blogged about a petition being circulated by the Association for Women in Mathematics (AWM) calling for the removal of Dr. Camilla Benbow from a panel created to investigate how to improve student performance in mathematics. The petition found its way to the Project NExT email list yesterday afternoon, and I sent in a reply message to the list referring them to the post for "a different take" on the situation.
That email was seen by many AWM people, including former AWM president Judith Roitman. She replied to me via email this morning, and I am posting the email here with her permission:
Benbow continues to cite the 13:1 ratio even though over the years it has dropped significantly — the latest figures are slightly under 3:1.
There are other methodological problems with her work that people in the field can tell you about, but our main concern is that she does not acknowledge a striking change in data. I.e., she goes around saying that things are the way they aren’t. This is not good science.
There are a lot of serious studies of girls and math coming to various conclusions — it’s a thriving field in education, ed psych, gender studies, etc.. AWM takes no stand on them.
I got feedback from other people saying basically the same thing. I can’t speak to the methodological merits or demerits of this research. There does appear to be substantive evidence that says that Benbow’s original results are not valid any more; and if a person sticks to invalid results despite mountains of evidence to the contrary, then I can certainly see a cause for concern.
But the AWM’s statements, particularly a lengthy statement from president-elect Cathy Kessel, seem to focus not so much on methodology as the premise of the research:
But our greatest concern is that its vice-chair, Dr. Camilla Benbow, is best known for the hypothesis that there are inevitable gender differences in favor of males at the highest level of mathematical performance. This hypothesis has already done serious damage (citations are below); furthermore, there is substantive evidence against it (again, citations are below).
We are concerned that Dr. Benbow is so closely identified with her 1983 statistics and hypothesis that her presence on the Panel signals – in perception or in reality – a bias against women and girls.
(All emphases added; whole thing here.) In other words, the greatest concern isn’t so much Benbow’s research per se as it is her affiliation with the results of that research; it has to do with what she is "best known for" and "closely identified with". But what if a person did research that was methodologically sound that came to similar conclusions as Benbow’s, and thereby became "known for" the same thing? What would happen if they got appointed to a national committee?
If research methodology is the problem, then the AWM needs to be up-front about that, by saying something to the effect of "Camilla Benbow is a bad scientist and is unfit to determine public policy based on her lack of expertise", and leave perceptions and affiliations out of it.
Powered by Qumana