Women also tend to be very influenced by images in the media, and there are not many favorable portrayals of scientists & engineers at all, let alone female scientists & engineers.

When the women wasted a year finding out that they hated engineering just as much as they expected, it turned out that the parents and guidance councilors weren’t doing them a favor at all.

I don’t know why everyone gets to excited over the fact that there are gender differences in cognition. Wouldn’t you EXPECT it, given all the other gender differences that exist?

Next thing we know we’ll be told how much better it is for the government to tell us what work to do because then there will be more women engineers.

As a woman engineer, I regret the choice! I loved getting my hands dirty on the machines, but once kids came along, I couldn’t support the 45+ hours needed plus the on-call type atmosphere when the manufacturing equipment malfunctioned on the weekend. Now I homeschool and do my own math curriculum because I don’t like anyone elses.

Yep. We are awash in scientific literature showing a cognitive gender gap in mathmatics. This mathematical gap increases gradually with age until puberty, when the gap explodes. And there are two features to this gap:

1) a higher male mean (smarter on average, men have more spatial ability, probably evolutionary from hunting)
2) a greater male variance (there are more dumb and smart men, nature can afford to take risks with men and not women)

While many studies are flawed due to sample size or whatnot, a good example here is the Project Talent study in 1960: 73,000 15 year-olds, both students and nonstudents given an all-day battery of 23 cognitive tests. Results: a mathmatical mean (male-female) difference of 0.12 standard deviation and a 1.20 (male/female) variance.

People forget that even a slight IQ gap in the middle of the bell curve translates to massive differences at the tails. By the time you get to the top slice of mathmatical nerdville (or even nerdville in any subject due to the variance issue) statistically we should expect few women. And this is what we see. This is the “Larry Summers problem” of high-end academia.

The studies say little about performance differences at the upper tail of the distribution which is especially relevant for becoming an academic in math or physics. The SAT does a poor job of distinguishing math talent for those who can score consistently around 760-800. And it is this tail which is relevant for the top ranks of academia. Consider that in the William Lowell Putnam Math contest, only a couple of women have ever been fellows. [Even the Elizabeth Putnam Fellowship -- which is only awarded to women, and not necessarily top scorers -- has only been awarded to 7 women]. One doesn’t have to be a math major to take the test and freshmen at the elites usually have enough math to handle it. But you have to be extraordinarily gifted to do well on it. Similarly only a few women are able to do well in Harvard’s Math 55 or the most hardcore courses at Caltech, Chicago, MIT or Princeton.

So I suspect that Rosenbloom’s work is very relevant for standard engineering and computer science. It does not add much to the “Larry Summers problem” in high-end academia.

I chose the computer industry as a post-military career, but not as a programmer/technology whiz - as an instructor. It gives me the best of both worlds - the geekiness that I love, but the people contact that I can’t live without. And none of the hassles that accompany the education industry.

I used to go to vendor technical training held in hotels, and I’d know I was in the right room when I saw a bunch of scruffy guys in shorts/t-shirts. I was typically the only female in the room. That’s a good 15 years ago - don’t know what it’s like now. I deal more with end-user training than the high-tech stuff, these days.

That’s because they’re smarter than we are.