Tuesday, July 13, 2010

SATs and PhDs

For better or for worse, the Tierney articles got me thinking: is it really true that those with SAT math scores at the 99.9 percentile level are that much more likely to get a doctorate in science or to win tenure at a top university than those at the 99.1 level, as Tierney claims? My gut feeling, from having interacted with hundreds of PhD astronomers, including myself, is that it isn't true. Heck, I doubt the scores themselves are even that accurate. Also, while I can see that math scores are important for achieving in science, verbal scores are surely very important also. After all, the way scientists communicate our ideas to each other is through talks and papers, both of which surely involve good verbal and communications skills. In which case, shouldn't girls' superior verbal scores balance things out?

So, if I were to conduct a thoroughly non-scientific internet survey on PhDs in astronomy and your SAT scores (if you can remember them) to analyze Tierney's thoroughly non-scientific claim, would you participate? If there's sufficient interest, I'll go and figure out how to set up a survey.

6 comments:

  1. sounds like fun! I might even dig up my SAT scores.

    Not that SAT scores would be not worth doing, but wasn't Tierney talking about PSAT scores of 8th graders? (I really don't want to re-read the article to check myself...)

    ReplyDelete
  2. I've always wondered how the astronomer population would compare to the mean, and how that evolves with career stage, and other things. I'd add mine.

    ReplyDelete
  3. I would participate. I'm curious to see the results and the scatter in the results.

    ReplyDelete
  4. A very good point!

    I think most of us would agree that success in astronomy is more about stubbornness than raw intelligence (however you measure it).

    I would participate.

    ReplyDelete
  5. Sure. I'd participate. I'm male, but also fall in the high verbal rather than high math category...

    ReplyDelete
  6. Tierney was referencing the SAT scores of 7th graders who took the SAT for a talent search in the 70s. If you look at the actual article that Tierney references (Wai, Lubinski, and Benbow, Journal of Educational Psychology, 2005, Vol. 97, No. 3, 484-492), the data for women seems incredibly incomplete, and based on small number statistics. The # of women with math-science doctorates in the bottom quartile of the 1972-74 sample is 0/151, and 2/130 in the top quartile; for the 76-79 sample it's 2/50 in both the bottom and top quartile. The male sample size is larger, so a trend there is more believable.

    I personally don't understand how Tierney continues to write for the NYT, which I find to otherwise be an astounding news source.

    ReplyDelete