My Fabulous Geek Career (O'ReillyNet)

Yes, it seems to be from table H-1 on page 177, "Computer and mathematical occupations". Page 180 puts female participation in "Engineering occupations" at 10.9%, lower but still dramatically higher than in FOSS.

>I suspect that those also include people in the Academia.

page 181: "Science and engineering occupations do not include postsecondary teachers." Not that it matters; the same report also states in table H-23, page 247, that the US has only 1030 female PhDs working at universities as "computer and information scientists", as compared to the >400,000 in the 28% number. (Also, why the heck do you care if it includes people in academia, presumably people in academia would be good candidates for FOSS participation too?)

It may include occupations other than computer programmer per se; it's possible the FLOSSPOLs people were a little hasty in calling that 28% the same as programmers per se, and I can't immediately find any way to get a more accurate number other than downloading the full US Census datasets and crunching them myself. However, note that page 180 puts female participation in "Engineering occupations" at 10.9% -- these are non-computer engineers, who I doubt have substantially *more* female participation than computer engineers (and this certainly does not include anything like graphic designers or what-have-you). Note that while lower than 28%, it is still dramatically larger than any numbers anyone gets for FOSS.

>And if you don't have real data, then where is thi data point of 1.5 % from?

I said I didn't have real data on job security. The 1.5% comes from... real data.

>It is from a different research.

Yes -- research performed in the EU, also replicated at Stanford (Stanford found 1.6%): http://www.stanford.edu/group/floss-us/report/FLOSS-US-Re...

If one is worried about bias from surveys (maybe women hate responding to them or something, not that there's any reason to think that), we can also look at the demographics of existing well-defined groups of FOSS developers.

Gentoo developers: "Less than 1% are female" (http://project.repogirl.net/doku.php) (IIRC they have 3, one of whom is transgendered and was male at the time he became a developer?)

Ubuntu seems to be ahead of the game with 2.4% (i.e., 7 total): http://www.eskar.dk/andreas/output/PersonalProfile.HTM

The Google Summer of Code mentors meeting for 2006 had approximately 100 attendees, of which 2 were female.

I'll also note that as far as I know, there are exactly zero women working on any of the FOSS projects I work on, out of some dozens of participants.

The 1-2% finding seems to be very robust.

>I just don't assume in advance that there is something that needs correcting. Is there something that needs correcting? please demonstrate.

Some of us find it pretty obvious without any numbers -- we find the blatant sexism displayed by many of the posters in this forum, for instance, to be a problem all by itself (or if you want numbers, note the in FLOSSPOLS ~75% of women respondents have observed discrimination against women -- though only ~20% of the men have, presumably because there were no women around for them to see being discriminated against). The implicit sexism in the posters who continually find new excuses each time their previous one is knocked down is also rather grating. But I've shown a whole pile of numbers now too, and every piece of data I can find gives percentages that are a factor of 10 lower than any comparable groups I can think of.