GENDER GENIE
Instapundit has revived interest in the Gender Genie, prompting Andrew Sullivan to boast of his manliness:
My average score is over 2-1 male, and the blog is the malest--much less metrosexual than even the Insta-man And I'm a big fag. It would indeed be interesting to see whether gay male writers end up being more or less reliably identified as male by this program. My own hunch suggests that gender is a far more profound determinant of human behavior than sexual orientation.
I hate to burst your bubble, Andrew, but it's not hard to write more manly prose than InstaPundit. Here are a sample scores from TSOS selections:
Female Score: 2479
Male Score: 3885Female Score: 934
Male Score: 1587Female Score: 1463
Male Score: 2525
And I'm a heterosexual woman with raging hormones and an evolutionary-psychology-approved waist-hip ratio of 0.70. I'm not masculine; I just like definite articles. I'm not saying gender is socially constructed, but I wouldn't recommend that anyone look for dates based on Gender Genie scores. It's not just me and not just opinion journalists, D Magazine's restaurant critic, Nancy Nichols, reports scoring 100% male, while executive editor Tim Rogers says, "Turns out, I'm a chick."
Reader Elf M. Sternberg writes:
The Gender Genie algorithm, which first appeared in the NY Times' "science" section, is a poor popularization of the algorithm as it appeared in the original academic literature. I have the original paper and that algorithm is meant to be applied to fiction; applied to non-fiction, the authors admit, the algorithm is no better than random chance at detecting an author's gender. A much better alogrithm, the one that has an "80%" chance of detecting author's gender correctly, needs to be taught on a large sample to generate a massive statistical measure of male vs. female characteristics in text.
Even applied to fiction, the popular algorithm is not much better. It seems to think I'm a woman, at least 97% of the time.