What to wear?

I saw an amazing -- as in amazingly inane, off-base, bizarre -- thing on the web recently. A woman wrote to a fashion adviser, asking for advice not for herself, but for the female colleague of her husband. The female colleague of the husband did not ask for or need advice, but the woman writing to the fashion adviser was concerned because her husband's colleague wasn't dressing well enough to succeed in her career.

This caught my attention because the apparently poorly dressed woman is in a job similar to mine - science professor. The ignorant fashion advisor confirmed that yes, in fact, the female science professor could be harming her career as a researcher by not dressing appropriately. She might not be productive enough at publishing and getting grants because she is not wearing the right clothes. There were links to sites where one could buy the *appropriate* attire for a female professor, but the recommended garments were hideous.

Does one's appearance matter in academia -- specifically in the corner of academic with which I am familiar: the physical sciences? There is no one answer to this because there are different phases and aspects of one's career to consider:

1 - grad students. It doesn't matter what you wear in grad school (unless you happen to have an advisor with a dress code), but it might matter what you wear to conferences or a job interview. In my field, only people with connections to industry, government, or administration wear anything resembling a business suit. When I was a student, I was advised to wear make-up and do something with my hair and dress up more or I would never get a job, but I ignored that advice, and here I am!

2 - professors

a - teaching. Some of my colleagues (male and female) have done "experiments" to test whether what they wear has any effect on their teaching evaluations. It turns out that when these colleagues have dressed well for every class in a semester, their teaching evaluations are higher than their usual average. It isn't a huge effect -- a lousy teacher isn't going to get great evaluations because of their wardrobe -- but my colleagues believe the effect is statistically significant. I tried dressing up early in my career just because I looked so young and had trouble convincing people (students, staff, other faculty) that I was a real professor. It didn't work very well. Students would come to my office and look around and say "Wow, it's so great that they give you an office even though you're not a professor." I would say "Why don't you think I'm a professor?" and they were startled and had no answer. Other times, I'd be sitting at my desk working in my very own office and someone who didn't know me would come in and (rudely) order me to tell Professor X (me!) something or give her something. They thought I was a secretary. I learned two things from this: (1) many people don't treat secretaries very politely, and (2) even today, people see a young woman and think "secretary" before they think "professor", even if there are visual clues to the contrary. Anyway, nowadays I don't dress up in any particular way to teach, but I don't wear T-shirts either, and I don't think my students or my career are harmed by my appearance.

b - research. How you dress has absolutely no effect on research, on grant success, on publication record. Some of the most talented women in my field dress abysmally; others always look fabulous and stylish. IT DOESN'T MATTER.