« September 2006 | Main | November 2006 »

October 30, 2006

Data: You Don't Have A Geek To Get It.

I am in data geek heaven. The National Assessment of Educational Progress, the Bureau of Labor Statistics, and the National Science Foundation all have easy to use tools to access data. Well ok maybe they aren't really easy to use; but they aren't hard either. With these tools you can break out their data by gender and by race/ethnicity. And they have a number of other variables you can use to slice and dice the data including age and level of education.

In the midst of all this good news, I have a piece of bad data news. It's the College Board. This week I was curious as to changes in the numbers and percentages of girls interested in going into the sciences so I went to my trusty source, the College Boards Annual Profile of College Bound Seniors. For the first time in at least ten years, the information on intended college major is not broken out by sex. While the federal government is moving forward allowing use to look at gender and race/ethnicity; the College Board is moving backwards and not allowing us to look at either when it comes to college majors. Please folks give us back the gender breakdown and if we are really nice maybe you could give it to us gender by race/ethnicity.

October 25, 2006

Journalists and Math Part II

October 8th I posted "A Perfect Storm of Misinformation" talking about Mark Liberman's suggestion that a discomfort with math among journalists might be one possible reason why so much research is reported so badly. This month's American Way, the in-flight magazine of American Airlines, made me think he is right. Read it and weep.

Someone named Barry wrote a letter castigating the editor for getting the year that Neil Armstrong walked on the moon wrong. Her response:

"Sure we remember, Barry, but that was 35 ... oh, wait ... 37 years ago. See? We’re still having trouble with numbers. Which is why we’re journalists instead of mathematicians. But thanks to you and everyone else for making us aware of our error.

October 20, 2006

Great Title, Good Content, But People Please Watch The Vocabulary.

FairerScience friend David Mortman recently sent us a link to Feminist Cyborgs:
Teaching like a Feminist in the Computer Classroom
. This transcript of a workshop is quite interesting, talking about issues related to feminist instruction/pedagogy as well as ways to use technology to create "different circuits of authority in the classroom” and to make assessments "on the basis of what students demonstrate they know and can do.” Those interested in instruction should take a look at it.

However it might be best if you skip the first page and a half. Sentences like:

"…critical teachers may find ways to sponsor infidel heteroglossia: multi-accentuated languages inside and outside of educational institutions that critique androcentric narratives or hegemonic
interpretations of everyday experience that deny the oppression of women and stifle hope for social change.”

scare me. They scare me a lot and make me worry if we are ever going to be able to effectively communicate our work to anyone other than each other.

October 17, 2006

Don't believe everything you read, even if it includes numbers

At the risk of branding FairerScience a fan site for Mark Liberman, I'd like to point you all to an article he wrote at the beginning of the month for The Boston Globe. In Sex on the brain, Liberman tackles the issue of urban legends that look like science.

This is one of our big frustrations here at FairerScience. We all know men and women are different in real and observable ways, and that makes it easy for us to believe it when people throw around numbers that seem to support and expand our beliefs. And most people aren't going to take the time to check citations and read the studies behind the story (if, in fact, there is a study, which there often isn't.)

Disappointingly, authors of popular science books and articles frequently use questionable studies or common knowledge to support their findings regarding sex differences. In this article, Liberman checks some of the numbers in the recent bestseller The Female Brain. Many of them, unsurprisingly, are unsubstantiated.

And, amusingly, as the friend who pointed this article out to me noted, the folks at the Globe who chose a graphic for the story obviously went with the headline rather than the content of the story, choosing a graphic that supports the urban legend numbers rather than the actual critique of such that Liberman offers.

October 13, 2006

Testosterone: boys will be boys?

We often hear about the effects, positive and negative, of testosterone. This "male" hormone (in scare quotes because both men's and women's bodies produce both estrogen and testosterone) is credited with enhanced libido, energy, and immune function and blamed for violent or antisocial behavior and baldness.

Testosterone is also the subject of a Yale School of Medicine finding, reported CNN: Too much testosterone kills brain cells.

I was interested to read this, and, I'll admit, I found my bias toward thinking that the natural system has it pretty well figured out was supported by Barbara Ehrlich's statement "Too little testosterone is bad, too much is bad but the right amount is perfect." It's not a deep observation, but it is satisfying.

Satisfying, that is, until I got to a later statement from Ehrlich, which sounds yawn-inducingly like yet another biological excuse for boys to be boys: "Next time a muscle-bound guy in a sports car cuts you off on the highway, don't get mad -- just take a deep breath and realize that it might not be his fault."

It may be satisfying to stereotype all muscular men as cavemen, and to blame bad behavior on hormones rather than plain old bad manners, but it sure isn't scientific.

October 08, 2006

A Perfect Storm of Misinformation

I was going to post about the snarky, funny and even a little sad science journalism quiz in yesterday's HeadsUp, but once I started reading Mark Liberman's response in the Language Log, I knew that was where FairerScience had to go.

Math phobia, Liberman, suggests is one possible reason why so much research is reported so badly. As he explains. "I do realize that some people freeze up and stop thinking whenever they see a mathematical symbol or term (even something routine like r2 or p<.05). In this state of intellectual desperation, if forced by circumstance to pretend to understand what's going on, they clutch reflexively whatever simple-minded description comes most quickly to hand."

He makes a lot of sense. If you only know what an average is, when confronted with an average for group A and a lower average for group B, it is tempting to conclude everyone in group A is better than those in group B. If someone says the results are "statistically significant" and you don't know that statistical significance doesn't mean that a result is meaningful, just that it's real; then it is tempting to conclude that group A is A LOT better than group B and so on.

I'm reminded of when I was an expert witness on the Citadel sex discrimination case (a story that will have to wait for another entry). At a court recess, the attorneys called me over saying they had a question for me. The question, "Pat exactly what is a standard deviation?"

In response to that question I wrote a short piece Overcoming Test Anxiety: Measurement and Statistics for Lawyers and Others. If you've ever wondered "exactly what a standard deviation is" or even if you haven't, feel free to read it and pass it on to anyone who is, or should be, reading and using research on gender and science.

Liberman concludes his entry thus: "Unfortunately, this ignorance and pretense combine with darker motives of sensationalism and pandering to stereotypes, creating a perfect storm of misinformation."

I agree, but if we are ever to have FairerScience, we've got to do more than ride out the storm; we have to change the conditions that created it.

October 03, 2006

The X-Gals

Now that is definitely a title. It is not one, however, that I would recommend googling on while at work. The results are definitely not work place friendly. The X-Gals themselves are. They are "a group of nine female biologists who began meeting weekly over a few beers in 2000" as several of them wrote up their dissertations. Their name is "a double-pun on the X-Men superheroes and on X-Gal, a laboratory chemical sometimes used in biology."

The X-Gals are writing a series on the personal and professional challenges of life in academic science beginning this week in the Chronicle of Higher Education.

Their start is depressingly accurate. "Women in biology earn roughly half of the Ph.D.'s awarded annually by American universities yet are progressively more underrepresented in postdoctoral, tenure-track, and tenured faculty positions. We all know that. We also know that women with families are less likely to earn tenure than our childless peers." I am hoping that they'll have some good news as well and lots of strategies to deal with the challenges.

Unfortunately on-line access is limited to Chronicle subscribers, but I am hoping that those of us without subscriptions have friends who will share.

October 01, 2006

Thank You Language Log

For several weeks I have been planning to write an entry about the Language Log, but other things have gotten in the way. Today I just have to do it; I owe them. Last night I knew that ABC's 20/20 was going to do a segment on sex differences and I knew that I should watch it. But I have watched them, particularly John Stossel, mangle science in so many earlier segments that I just didn't have the heart to do it. Being made of sterner stuff than I, the Language Log did.

So please, please, please read about how they went into the real research to debunk some of the "some striking fragments of misinterpreted science" that made it into the "Let Science Decide" segment on last night's 20-20 including this one: "The male brain … actually has a harder time processing the female voice versus the male voice, which is a possible explanation to why we don't listen when our wives call us."

Also consider reading the Language Log on a regular basis. One of the goals of FairerScience is to help researchers in gender and the sciences better communicate their work to the media, policy makers, and advocates while helping media to better understand issues of gender and science, technology, engineering and mathematics (STEM). Reading the Language Log has been a great (and entertaining) way to learn more about what we have been doing wrong. Who could have figured a numbers geek like me could learn so much from linguists?