Following Glenn Reynolds’ example on Instapundit, I put a 609-word blog entry of mine through the GENDER GENIE. The results were:

Female Score: 341
Male Score: 1521

I guess I’m not surprised by this. I’m frequently mistaken for male on the internet, and was accused of being a bulldyke in high school (and assumed to be one in college, although in college it’s a positive thing). And with hobbies like cars, guns, military strategy, and an interest in pornography, one might be forgiven for making the assumption. Even my mother thought I was going to be a boy before I was born.

I am, however, heterosexual and happily married. Go figure.

I tried a different blog entry to see if I could get a different statistical result, this time with a 568-word entry.

Female: 1245
Male: 1425

Now that’s a little more like it. Although, an examination of the word table results indicates that the majority of the “female” score was derived from 6 instances of the word “him”, which is to be expected in a political post about Grey Davis.

Now, shoving through a passage out of one of my works of fanfic (1397-word chunk) yielded this:

Female: 2558
Male: 1783

I dunno. Is there really something to this? The authors of the Gender Genie say it should be accurate approximately 80% of the time, based on occurrences of specific words that have a “male” or “female” value.

Personally this looks kinda like a simple parlour trick. Not sure if there is any relevant psychological conclusion to be drawn from this.

2 Responses to “I'm One Butch Chick”
  1. shell says:

    I am female and my blog partner is male. I ran randomly chosen posts of each of us through that and both were wrong. It’s not even a parlor trick. It doesn’t even pretend to work.

  2. Kyle Haight says:

    Just for kicks I dug up a letter I wrote a decade ago, to a girl I knew in college. Based on the Gender Genie score (1437 female, 1545 male) I think I must be metrosexual. Some shorter text samples I tried actually came out female, but I’ll cut them some slack since they were under 500 words and they explicitly state that the accuracy of the algorithm degrades on samples that small.

    Still, I don’t think there’s anything really amazing going on here.

Leave a Reply