Friday, February 25, 2011

Is Discrimination Largely a Thing of the Past?

In the February 11, 2011 issue of AASWomen, Dr. Abigail Stewart of the University of Michigan provided comments on an article, "Understanding Current Causes of Women's Underrepresentation in Science" by Stephen J. Ceci and Wendy M. Williams. Dr. Stewart finds that these authors use their review of existing data on hiring, publishing, and funding in science research to conclude that discrimination does not exist.

Nancy Brickhouse (Associate Director Harvard-Smithsonian Center for Astrophysics and Division Head, Solar, Stellar, and Planetary Sciences) and Andrea Dupree (Senior Astrophysicist at the Smithsonian Astrophysical Observatory) bring us their thoughts on the Ceci and Williams article:

1. As Abby Stewart pointed out, there is no new data here. This is just a meta-analysis.

2. Most if not all of the studies used in the article are from the life and behavioral sciences, which are in general less math-intensive than astronomy and physics and also have better representation of women (e.g. Journal of Biogeography, Behavioral Ecology, Nature Neuroscience).

3. The article cites studies from NSF, NIH, and the Department of Agriculture. The latter 2 agencies do not primarily do math-intensive research. The percentage of math-intensive research even at the NSF is about 20% (judging by the FY2012 Budget request at the Division of the Mathematical and Physical Sciences MPS). Thus these studies are not relevant.

4. The examples from outside the US used to document their case are not relevant, since cultures, funding, child care, health care, etc are all different as well.

5. We agree with what Abby Steward wrote in AASWomen "They [reach their conclusions] mainly by focusing on whether similarly-situated men and women scientists have similar outcomes. The answer to that is yes--and that is indeed good, and perhaps not surprising. But the problem we have is actually that men and women scientists are NOT similarly situated--a point they note, but that is mostly overlooked in media accounts, perhaps because less ink is shed on it in the article."

6. Many of the set-out quotes providing "expert documentation" in the article are from other papers by the same authors.

7. Their attribution of differences in outcomes primarily to "choices" (freely made or constrained) appears to be the ticket to media attention, no surprise that it resonated with John Tierney of the NY Times Science section.

8. If discrimination no longer affects peer review for proposals and publications or hiring decisions (and we do not feel they have demonstrated this for the physical sciences), that is certainly a good thing. It means that the formal structures are working at the agencies responsible for peer reviews, at the editorial boards, and at institutions that are hiring (through Human Resource departments or oversight at the Dean's level). But there are many other forms of discrimination not addressed through formal structures. What happens in the research group meeting, between students and advisors, and in other informal work settings can have a big impact on the work climate and advancement, and this is the area where gender equity studies at some institutions (e.g. here at the CfA) show there are still many problems. See:

http://www.cfa.harvard.edu/do/geneq

9. Abby Stewart also wrote that the article points to "institutional barriers... that are worth attention." We agree.

1 comment :

Sarah said...

Yeah, John Tierney and some other commenters really misinterpreted this paper. A far better overview is given in Slate.