10 ms·

Why Men Don’t Believe the Data on Gender Bias in Science

5 points by alannallama 9y ago

mpweiher 9y ago
Because the data shows something else entirely and is always misrepresented. For example, the lab manager study that's quoted was for a lab manager in psychology. Which actually has greater female than male representation. Minor detail. Oh, and the effect could only be reproduced if the fake applicants were both (a) mediocre and (b) exactly equally qualified. Whereas Ceci and Williams found a 2:1 hiring advantage for women in STEM, with a much larger and to the best of my knowledge also much more careful study. Facts. Matter.
- pron 9y ago
  > For example, the lab manager study that's quoted was for a lab manager in psychology. Actually it wasn't. Quote, "A broad, nationwide sample of biology, chemistry, and physics professors (n = 127) evaluated the application materials of an undergraduate science student who had ostensibly applied for a science laboratory manager position." > Which actually has greater female than male representation. Minor detail. It is a minor detail, even if the study really were about psychology lab assistants, which it isn't, because the hypothesis isn't about total participation but about bias. Participation numbers alone don't tell you about bias. That's like Tucker Carlson, who recently asked how NFL owners can be racist if 70% of their players are black. > Facts. Matter. But George Washington was the first president of the United States! Facts. Matter. Only, it's kind of hard to see how they matter unless you understand the issue you're commenting on and the hypothesis those facts ostensibly support. You seem to be under the impression that the problem is simply the number of women in various arbitrary professions. I understand how someone would come to that conclusion if they only read headlines. It isn't the issue, however, it isn't the hypothesis, and therefore you'll keep bringing up facts that don't really matter for the subject that's actually discussed, rather than the one you think is being discussed. This is because you don't really care about what the issue is; you just decided it was wrong based on some gut feeling you may have.
  - mpweiher 9y ago
    > Participation numbers alone don't tell you about bias Well, if "bias is the answer, what was the question?" is your "scientific" approach, then they obviously don't matter. However, it turns out that neither chemistry nor physics are "tech" and the gender ratios in the physical sciences are fairly even (even for chemistry, slight male skew for physics last I checked). Add in psychology, and using this study as a proxy for engineering or tech is just malpractice. Which means par for the course for gender "studies". Anyway, it looks like people have a bias for the underrepresented gender in a particular field, as the following study shows: "The team found a bias in favor of women in math, physics, and philosophy, which increased with the degree of male domination in the field" "They found about a 10 percent bias in favor of female test takers in the non-gender–blind oral tests in math, physics, and philosophy. The opposite was true for fields where men are traditionally underrepresented: The team found a 3 percent to 5 percent bias in favor of male oral test takers in literature and foreign languages" http://www.the-scientist.com/?articles.view/articleNo/46681/title/Slight-Female-Bias-in-French-Science-Teacher-Exams/ http://www.the-scientist.com/?articles.view/articleNo/46681/... Which is kind of what you would expect humans to do (slightly favor the group that is underrepresented), and surprise, that is in fact what humans do. And when the underrepresentation is stronger, the bias for the underrepresented population increases, see the 2:1 hiring advantage for for women for tenure track positions in STEM found by Ceci/Wiliams. Stronger effect, much larger sample, more careful study, much less quoted than the "lab manager" one, which to remind everyone only saw an effect for mediocre candidates (so people you didn't want to hire in the first place) that were calibrated to be exactly equally qualified.
    - pron 9y ago
      > However, it turns out that neither chemistry nor physics are "tech" and the gender ratios in the physical sciences are fairly even (even for chemistry, slight male skew for physics last I checked). I don't understand what it is that you're arguing. > And when the underrepresentation is stronger Again, I don't think you understand what the topic is, so I don't see the relevance of your citations. What kind of bias do you think we're talking about? Even though I hate doing this (b/c if you haven't bothered clicking through the links I posted, then you're not interested in the subject), but just to break the monotony of repeatedly pointing out the absurdity of debating a subject you know little about and wish to learn of even less, let me give the elevator-pitch summary: It is not about participation in arbitrary professions; it is not about bias in tests; it is not about biological differences; it is about the easily verified fact that women have been marginalized from positions of power[1] at least for centuries. Test scores have little to do with power; Participation in being psychology lab assistants has little to do with power; bias in promotion in certain power-conferring occupation has a lot to do with power. Studies in the field may, therefore, focus on two things: either on the status or process of said power imbalance, or on the mechanism of its creation. The paper which is the focus of this article (the last one), is of the second kind. Next you're going to show me research showing bias in favor of women when it comes to smiling. That may be nice, but is not relevant to the discussion. If you show me research showing how socitey confers more power on women than men, then that would be something to consider. [1]: https://en.wikipedia.org/wiki/Power_(social_and_political) https://en.wikipedia.org/wiki/Power_(social_and_political)
gamechangr 9y ago
"One paper found ....instead the male faculty favored fake research". Who cares what one paper found? That's not proof. There are 10,000 papers on both sides.
belorn 9y ago
> These lawsuits highlight the real reason for the lack of women in science In an article about science, a claim like that without strong support is in very bad taste. If you want to prove causation, you should at least first prove correlation. If we measure how much an industry believes that a specific gender is suitable for the job, do we get a correlated number for gender segregation? For an assertive statement that we have now found a scientific proof for the real underlying reason for gender segregation, such question should be easily answered.

5 points by alannallama 9y ago

Because the data shows something else entirely and is always misrepresented. For example, the lab manager study that's quoted was for a lab manager in psychology. Which actually has greater female than male representation. Minor detail. Oh, and the effect could only be reproduced if the fake applicants were both (a) mediocre and (b) exactly equally qualified. Whereas Ceci and Williams found a 2:1 hiring advantage for women in STEM, with a much larger and to the best of my knowledge also much more careful study. Facts. Matter.

pron 9y ago
> For example, the lab manager study that's quoted was for a lab manager in psychology. Actually it wasn't. Quote, "A broad, nationwide sample of biology, chemistry, and physics professors (n = 127) evaluated the application materials of an undergraduate science student who had ostensibly applied for a science laboratory manager position." > Which actually has greater female than male representation. Minor detail. It is a minor detail, even if the study really were about psychology lab assistants, which it isn't, because the hypothesis isn't about total participation but about bias. Participation numbers alone don't tell you about bias. That's like Tucker Carlson, who recently asked how NFL owners can be racist if 70% of their players are black. > Facts. Matter. But George Washington was the first president of the United States! Facts. Matter. Only, it's kind of hard to see how they matter unless you understand the issue you're commenting on and the hypothesis those facts ostensibly support. You seem to be under the impression that the problem is simply the number of women in various arbitrary professions. I understand how someone would come to that conclusion if they only read headlines. It isn't the issue, however, it isn't the hypothesis, and therefore you'll keep bringing up facts that don't really matter for the subject that's actually discussed, rather than the one you think is being discussed. This is because you don't really care about what the issue is; you just decided it was wrong based on some gut feeling you may have.
- mpweiher 9y ago
  > Participation numbers alone don't tell you about bias Well, if "bias is the answer, what was the question?" is your "scientific" approach, then they obviously don't matter. However, it turns out that neither chemistry nor physics are "tech" and the gender ratios in the physical sciences are fairly even (even for chemistry, slight male skew for physics last I checked). Add in psychology, and using this study as a proxy for engineering or tech is just malpractice. Which means par for the course for gender "studies". Anyway, it looks like people have a bias for the underrepresented gender in a particular field, as the following study shows: "The team found a bias in favor of women in math, physics, and philosophy, which increased with the degree of male domination in the field" "They found about a 10 percent bias in favor of female test takers in the non-gender–blind oral tests in math, physics, and philosophy. The opposite was true for fields where men are traditionally underrepresented: The team found a 3 percent to 5 percent bias in favor of male oral test takers in literature and foreign languages" http://www.the-scientist.com/?articles.view/articleNo/46681/title/Slight-Female-Bias-in-French-Science-Teacher-Exams/ http://www.the-scientist.com/?articles.view/articleNo/46681/... Which is kind of what you would expect humans to do (slightly favor the group that is underrepresented), and surprise, that is in fact what humans do. And when the underrepresentation is stronger, the bias for the underrepresented population increases, see the 2:1 hiring advantage for for women for tenure track positions in STEM found by Ceci/Wiliams. Stronger effect, much larger sample, more careful study, much less quoted than the "lab manager" one, which to remind everyone only saw an effect for mediocre candidates (so people you didn't want to hire in the first place) that were calibrated to be exactly equally qualified.
  - pron 9y ago
    > However, it turns out that neither chemistry nor physics are "tech" and the gender ratios in the physical sciences are fairly even (even for chemistry, slight male skew for physics last I checked). I don't understand what it is that you're arguing. > And when the underrepresentation is stronger Again, I don't think you understand what the topic is, so I don't see the relevance of your citations. What kind of bias do you think we're talking about? Even though I hate doing this (b/c if you haven't bothered clicking through the links I posted, then you're not interested in the subject), but just to break the monotony of repeatedly pointing out the absurdity of debating a subject you know little about and wish to learn of even less, let me give the elevator-pitch summary: It is not about participation in arbitrary professions; it is not about bias in tests; it is not about biological differences; it is about the easily verified fact that women have been marginalized from positions of power[1] at least for centuries. Test scores have little to do with power; Participation in being psychology lab assistants has little to do with power; bias in promotion in certain power-conferring occupation has a lot to do with power. Studies in the field may, therefore, focus on two things: either on the status or process of said power imbalance, or on the mechanism of its creation. The paper which is the focus of this article (the last one), is of the second kind. Next you're going to show me research showing bias in favor of women when it comes to smiling. That may be nice, but is not relevant to the discussion. If you show me research showing how socitey confers more power on women than men, then that would be something to consider. [1]: https://en.wikipedia.org/wiki/Power_(social_and_political) https://en.wikipedia.org/wiki/Power_(social_and_political)

gamechangr 9y ago

"One paper found ....instead the male faculty favored fake research". Who cares what one paper found? That's not proof. There are 10,000 papers on both sides.

belorn 9y ago

> These lawsuits highlight the real reason for the lack of women in science In an article about science, a claim like that without strong support is in very bad taste. If you want to prove causation, you should at least first prove correlation. If we measure how much an industry believes that a specific gender is suitable for the job, do we get a correlated number for gender segregation? For an assertive statement that we have now found a scientific proof for the real underlying reason for gender segregation, such question should be easily answered.