3 ms·

Every field you mentioned had the same male majority, yet they all failed to keep women out. Well, almost every field you mentioned - actuaries were quite succe

Every field you mentioned had the same male majority, yet they all failed to keep women out. Well, almost every field you mentioned - actuaries were quite successful. http://www.prb.org/pdf08/scientistprofiles/occprof08_actuaries.pdf http://www.prb.org/pdf08/scientistprofiles/occprof08_actuari... If you want to appeal to "bog-standard human behavior", it would help if you showed that computing, physics and actuary have "bog-standard human behavior" while law and medicine don't. As I said, not even correlation has been demonstrated. Just so that you can understand what I mean by "correlations", let me demonstrate with a different theory. Theory: women are less likely to have high quantitative ability, so fields with higher quantitative ability requirements will have fewer women. Demonstrating a correlation would involve showing a downward sloping graph with quant ability requirements on the X-axis and %women on the Y-axis. Like the graphs here: http://slatestarcodex.com/2015/01/24/perceptions-of-required-ability-act-as-a-proxy-for-actual-required-ability-in-explaining-the-gender-gap/ http://slatestarcodex.com/2015/01/24/perceptions-of-required... Then I'd need to present evidence that women are less likely to have high quant abilities. Like in these papers here (note that the abstracts have been creatively worded to be published, skip to the data tables): http://www.sciencemag.org/content/321/5888/494.summary http://www.sciencemag.org/content/321/5888/494.summary http://www.jstor.org/discover/10.2307/2889145?uid=3738256&uid=2&uid=4&sid=21105200582321 http://www.jstor.org/discover/10.2307/2889145?uid=3738256&ui... http://www.ncbi.nlm.nih.gov/pubmed/21038941 http://www.ncbi.nlm.nih.gov/pubmed/21038941 http://www.ams.org/notices/201201/rtx120100010p.pdf http://www.ams.org/notices/201201/rtx120100010p.pdf Simply pointing out that humans act like humans fails - it predicts flat graphs rather than downward sloping ones.

tptacek 12y ago
I said extraordinary claims require extraordinary evidence. You just made two different extraordinary claims: that women (as a demographic) intrinsically lack quantitative ability, so much so that they're at a disadvantage in industry (not merely at a disadvantage in the competition for Fields Medals), and that computer science as an industry is uniquely dependent on quantitative ability, moreso than actuaries (where there is significantly greater female participation.) Your extraordinary evidence for this is a blog post with regression lines. Sometimes your rhetorical strategy on HN is bracing and useful. Like all message boards, HN is prone to flights to abstraction and goalpost moving, and the introduction of facts can cut through that and ground the discussion. Other times, it seems like you simply have far too much confidence in the significance of the facts you choose, and your overconfidence hamstrings you. I don't have to care whether male quantitative ability has more variance than female ability. I can sort of enjoy reading about it on Marginal Revolution without incorporating any of those conclusions into my identity or philosophy. I do care about the idea that results about variance in quantitative ability speak to CS ability. For the great majority of specialties in CS, they do not, at all. I think you may be blinded to this by the specialty you've chosen to work in, which is math-intensive. I'm a systems programmer --- in fact, a crypto-intensive systems programmer! --- and I will go to the mat against the idea that computer science requires atypical quantitative ability. At any rate: it is self-evident that professional software development doesn't require the quantitative aptitude required to compete for university math tenure.
- yummyfajitas 12y ago
  I gather you assign a lower prior to cognitive variations - that's fine. No point debating that at the moment, it's just an example to show what evidence might look like: a downward sloping graph with your predictor on the X-axis and %female on the Y. You don't have that graph. All you have is vague theories like "bog-standard human behavior prevents women from entering a field". This theory predicts that every historically male field should remain historically male, which is false.
  - tptacek 12y ago
    Every historically male field will remain male in the absence of a focused effort to improve balance, which is what happened in the demanding (and often quantitatively rigorous) professions that improved women's participation, and what's being contested in computer science. I am not making an extraordinary claim. I am looking at the numbers for female participation in CS, noticing that they are anomalous, all the more so given the relative size of the CS professions (they are enormous, which should have helped smooth out the anomaly), noting the incentives for participation (CS professions are, relative to the effort it takes to enter the field, well compensated, flexible, geographically well distributed, and not physically demanding) and observing that there are only two logical explanations: (a) Social and process factors are having the effect of encouraging male participation and discouraging female participation, or (b) Women are intrinsically sharply less capable of performing in software development despite a demonstrated ability to perform in law, accounting, medicine, and even other STEM fields. In the absence of something much more compelling than a blog post on slatestarcodex, I feel pretty comfortable ruling out (b).
    - yummyfajitas 12y ago
      So where is your evidence that quantitative fields did not have this "focused effort", but non-quantitative fields did? Can you present some evidence of focused effort in medicine, law and biology but none in computing and aerospace engineering? I.e., do you have an upward sloping graph with focused effort on the X and %female on the Y?

tptacek 12y ago

I said extraordinary claims require extraordinary evidence. You just made two different extraordinary claims: that women (as a demographic) intrinsically lack quantitative ability, so much so that they're at a disadvantage in industry (not merely at a disadvantage in the competition for Fields Medals), and that computer science as an industry is uniquely dependent on quantitative ability, moreso than actuaries (where there is significantly greater female participation.) Your extraordinary evidence for this is a blog post with regression lines. Sometimes your rhetorical strategy on HN is bracing and useful. Like all message boards, HN is prone to flights to abstraction and goalpost moving, and the introduction of facts can cut through that and ground the discussion. Other times, it seems like you simply have far too much confidence in the significance of the facts you choose, and your overconfidence hamstrings you. I don't have to care whether male quantitative ability has more variance than female ability. I can sort of enjoy reading about it on Marginal Revolution without incorporating any of those conclusions into my identity or philosophy. I do care about the idea that results about variance in quantitative ability speak to CS ability. For the great majority of specialties in CS, they do not, at all. I think you may be blinded to this by the specialty you've chosen to work in, which is math-intensive. I'm a systems programmer --- in fact, a crypto-intensive systems programmer! --- and I will go to the mat against the idea that computer science requires atypical quantitative ability. At any rate: it is self-evident that professional software development doesn't require the quantitative aptitude required to compete for university math tenure.

yummyfajitas 12y ago
I gather you assign a lower prior to cognitive variations - that's fine. No point debating that at the moment, it's just an example to show what evidence might look like: a downward sloping graph with your predictor on the X-axis and %female on the Y. You don't have that graph. All you have is vague theories like "bog-standard human behavior prevents women from entering a field". This theory predicts that every historically male field should remain historically male, which is false.
- tptacek 12y ago
  Every historically male field will remain male in the absence of a focused effort to improve balance, which is what happened in the demanding (and often quantitatively rigorous) professions that improved women's participation, and what's being contested in computer science. I am not making an extraordinary claim. I am looking at the numbers for female participation in CS, noticing that they are anomalous, all the more so given the relative size of the CS professions (they are enormous, which should have helped smooth out the anomaly), noting the incentives for participation (CS professions are, relative to the effort it takes to enter the field, well compensated, flexible, geographically well distributed, and not physically demanding) and observing that there are only two logical explanations: (a) Social and process factors are having the effect of encouraging male participation and discouraging female participation, or (b) Women are intrinsically sharply less capable of performing in software development despite a demonstrated ability to perform in law, accounting, medicine, and even other STEM fields. In the absence of something much more compelling than a blog post on slatestarcodex, I feel pretty comfortable ruling out (b).
  - yummyfajitas 12y ago
    So where is your evidence that quantitative fields did not have this "focused effort", but non-quantitative fields did? Can you present some evidence of focused effort in medicine, law and biology but none in computing and aerospace engineering? I.e., do you have an upward sloping graph with focused effort on the X and %female on the Y?