Wednesday, January 10, 2007

I'm less than half the man you are...

Great post at MetaFilter. It's particularly about women in science but illuminates the consequences of thinking in sweeping generalisations.

The most salutory of the articles linked to is the story of Ben (ne Barbara) Barres.

For a drier but more "evidence based" view see instead the story about impact factors. Women in science need to be 2.5 times more productive to be adjudged equal to their male peers.


Martin Sewell said...

Barres' thinking is absurd; of course people see and treat men and women differently, this is because men and women are radically different. The case of Norah Vincent is more revealing.

Wennerås and Wold (1997) seem to be oblivious of the risks of datamining, yet remarkably adept at presenting their statistics in a misleading manner.

Auntie Em said...

I'm not sure what you find absurd about Barres' thinking? That his work was considered better after becoming a man than it was when he was though to be a woman? Despite the fact that it was the exact same work?

Norah Vincent's story is a nice collection of anecdotes - but the women she's been dating don't sound like any of my female friends. I'm not saying such women don't exist but I know plenty of them who are self assures and intelligent enough not to need emailing every half hour, and who don't begin each relationship with the warning "where my ex went wrong". And I know a few needy men.

Ben Barras was that he was the same person as a man and as a woman, doing the same work, and yet was treated differently. Why? Because he was radically different his 'sister'? No - because lazy thinkers go: "Barbara Barras comes from the set of all women. Not many women have good careers in science. Therefore Barbara Barras is not good at science.

If you can't spot the fallacy there then it explains your confident statement that "Of course people see and treat men and women differently, this is because men and women are radically different". We're a pretty heterogeneous group and so are men.

Wenneras and Wold do use multiple regression to try to find the major factors, but they don't begin by data mining. They find the effect first then try to explain it. As it says in the paper (and on the blog):

"the peer reviewers gave female applicants lower scores than male applicants who displayed the same level of scientific productivity. In fact, the most productive group of female applicants, containing those with 100 total impact points or more, was the only group of women judged to be as competent as the least productive group of male applicants (the one whose members had fewer than 20 total impact points)."

I don't see how anything could be less misleading.

Martin Sewell said...

My point about datamining in Wennerås and Wold (1997) is where they write "Three out of the six productivity variables generated statistically significant models capable of predicting the competence scores the applicants were awarded..." and then proceed to choose the best and ignore the other three. My point about the presentation of results is that the answer to the question, "For a given "Total impact", how much higher, on average, is the men's 'Competence' score?" is "about 10%". This is hardly headline grabbing. Also, far more of the men had a medical degree, this may have been relevant. The paper may well have identified a "rule of thumb" generalization issue. This is normally swamped by the prejudice in all societies that is directed principally not towards women but towards lower status (that is, the majority of) men.

Auntie Em said...

the prejudice in all societies that is directed principally not towards women but towards lower status (that is, the majority of) men

"Women have higher poverty rates than men in almost all societies (Casper et al., 1994). In this paper, we compare modern nations on this dimension. We use the Luxembourg Income Study (LIS) to compare women's and men's poverty rates in eight Western industrialized countries circa the early 1990s: the United States, Australia, Canada, France, West Germany, the Netherlands, Sweden, and the United Kingdom. We define individuals to be in poverty if they live in households with incomes below half the median for their nation."

Percentage of single parent households headed by a woman in the uk: 93%

In a recent study, 81% of domestic violence victims were women attacked by men; 8% were men attacked by women; 4% were women attacked by women; 7% were men attacked by men.

Percentage of children in fatherless families in the UK: 20.7%

And now I'm tired.

Martin Sewell said...

* Women can charge money for sex, whilst men can not.
* Women are more than twice as likely as men to initiate domestic violence and more than twice as likely to use weapons in domestic violence.
* Men are sixteen times more likely than women to go to prison.
* More than four times as many men as women commit suicide.
* Men are ridiculed for impotence or “premature ejaculation” (sic), whilst if a woman fails to orgasm, it’s still considered the man’s problem.
* Female circumcision is considered a serious crime, whilst male circumcision is considered a joke.
* Riach and Rich (2006) made bogus job applications for analyst programmer vacancies in central London that were identical but for the sex of the applicant. They found direct sex discrimination against men to the extent that men are rejected compared to women at the application (i.e. before interview) stage at the rate of four-to-one.

Anonymous said...

Oh please - female "circumcision" is analogous to having your entire penis cut off, since the clitoris itself is removed. That is the anatomically analogous body part. These "facts" of yours, Martin, read as if you received them in a violently foaming googlegroups missive from an angry fathers group.

Post a Comment