I think it's safe to say that those of us who saw the caption "The Myth of the Asian Fetish" on the front page (which will be gone in a few days or less) have been had. It's worth noting here that titles in Slate often (always?) contain puns, offensive references, or misleading nonsense designed to draw you in rather than accurately describe anything.

Several earlier posters have already expressed doubt that the so-called "Asian fetish" had been debunked by the study (PDF, via the article). There are sound reasons for such doubt that have not yet been discussed.

A very small point: The study claims (in footnote 5) that 202 men participated, but (in table I) the numbers only add up to 201 (not significant, but still true). There were 200 women (from table I again), and nowhere is it claimed otherwise. Thus, we will assume that there were 401 total participants.

It's difficult to tell what's going on in the first place because of the way the study plays with numbers. For example, in table IIa, there are four races listed: White, Black, Hispanic, and Asian. The total number of subjects of those races is 348. One assumes that the discrepancy is due to the remainder being of races other than those four. The table claims that the 228 White participants make up 65.52% of the 348 subjects in those four races, which is true. However, it clearly does not make up 65.52% of the total number of participants. The table does us the favor of comparing apples to apples when it gives us the percentages of each race present in the Columbia graduate student population: There are 3978 White students out of 5793 total students of the four races previously listed, which it correctly restates as 68.67%. But there are far more than 5793 students of all races; the same table claims there are 10,569 students in four fields of study, and then breaks down by field and compares the percentages to those of study participants. The point after all of this is that table IIa appears to be completely worthless, never mind its relevance to the rest of the study. (Besides, I'm quite sure someone will tell me if I have screwed this all up.)

If we pretend we're certain that there were 401 total participants, 67 Asians would be 16.71%. As others have said, this vastly skews the results since only 3.64% of the U.S. population is Asian (from the 2000 U.S. Census, PDF). Using the percentage of women (50.9%) in the total U.S. population as a guide, Asian women represent 1.85% of all people in the U.S. and 8.48% (34) of the study participants.

There were participants who had grown up all over the world, including Western Europe (where there are plenty of White men). There may not be many Asian women in Western Europe; I don't know. Some of the White male subjects probably grew up in parts of the U.S. with few Asian women. It could well be that several White male subjects had Asian fetishes, but contact with Asian women on a daily basis (at Columbia) put them off it, or merely satiated it. Or the particular Asian women in the study weren't to their liking. Or they didn't want to choose Asian women every time because of how it would look. Or some extremely attractive women of other races chose to take part in the study, making the Asian women seem plain by comparison. Or the White males in the study liked their prostitutes Asian, their mistresses White, and their wives Hispanic or Black.

Who knows? In any case, what can be extrapolated from this ridiculous study of 401 graduate students at a single university, anyway?

P.S. Obviously, the above list of personality quirks is by no means exhaustive. These and many other factors affecting individuals and/or subgroups in the study are also just as likely to explain away any other conclusion (i.e. other than the "debunking of the Asian fetish") that was drawn by the study's authors.

The writers have to put something out to get paid.

Doesn't sem to matter how it smells.