There's some controversy over whether Amanda Doerty is really a hot chick or just a man pretending to be a hot chick. There's only one way to know for sure!
No, not by talking to her or playing doctor... by using the Gender Genie to analyze her writing style! I'll just paste in some text from her front page and...
Words: 9401
(NOTE: The genie works best on texts of more than 500 words.)
Female Score: 15607
Male Score: 21269
The Gender Genie thinks the author of this passage is: male!
Here are the details:
| Analysis | |||||||||||
| Feminine Keywords | Masculine Keywords | ||||||||||
| [him] | 3 | x | 73 | = | 219 | [some] | 33 | x | 58 | = | 1914 |
| [so] | 82 | x | 64 | = | 5248 | [this] | 36 | x | 44 | = | 1584 |
| [because] | 21 | x | 55 | = | 1155 | [as] | 50 | x | 37 | = | 1850 |
| [actually] | 6 | x | 49 | = | 294 | [now] | 21 | x | 33 | = | 693 |
| [everything] | 2 | x | 44 | = | 88 | [good] | 28 | x | 31 | = | 868 |
| [but] | 74 | x | 43 | = | 3182 | [something] | 18 | x | 26 | = | 468 |
| [like] | 21 | x | 43 | = | 903 | [if] | 94 | x | 25 | = | 2350 |
| [am] | 7 | x | 42 | = | 294 | [ever] | 5 | x | 21 | = | 105 |
| [more] | 34 | x | 41 | = | 1394 | [is] | 185 | x | 19 | = | 3515 |
| [out] | 21 | x | 39 | = | 819 | [the] | 376 | x | 17 | = | 6392 |
| [too] | 22 | x | 38 | = | 836 | [well] | 14 | x | 15 | = | 210 |
| [has] | 25 | x | 33 | = | 825 | [in] | 132 | x | 10 | = | 1320 |
| [since] | x | 25 | = | 350 | |||||||
So there you have it! Hot Abercrombie Chick is a man, man. Absolute proof, brought to you by Modern Technology.
Ok, not really "absolute", but the Gender Genie is based on research by some Israeli linguists (as described in this Nature article) and is around 80% accurate. I used a large sample of "Doerty's" writing and the gender skew is high enough that we can be very confident that the author is, in fact, a male.
Update:
Duh! Look, I hate to be a cynical misogynist, but girls that hot simply aren't that smart and that eager to discuss philosophy. I know lots of hot girls, and even the intelligent ones get by mainly on their looks. Why? It's easier, and all humans, male and female, generally follow the path of least resistance.
I knew it! I used your blog for this week and ...
Words: 1847
(NOTE: The genie works best on texts of more than 500 words.)
Female Score: 2809
Male Score: 2789
The Gender Genie thinks the author of this passage is: female!
LT beat me to it. Curses! Foiled again.
You say "girls that hot simply aren't that smart and that eager to discuss philosophy" and I guess that *in general* you are correct, although there are always one or two exceptions.
The Beautiful People are treated well from an early age and don't have to work hard to stand out or please their parents. This means that in later life most of them wouldn't be capable of standing out intellectually even if they wanted to.
Still, there are days when I can't help wondering whether my life would be more fun if I looked like a male model :)
I don't think your Gender Genie knows very much. I put in a couple of my recent blog posts. The one on science and culture indicated I was male and the one on ADD indicated I was female. I'm actually female. Probably not "hot" by many people's standards (being third-trimester pregnant with Kid #3 doesn't increase pulchritude much), but definitely female. Still, many people have mistaken me for a male based on my writing, so it doesn't surprise me.
My daughter, however, is strikingly beautiful... and thinks just like her mother. We'll have to wait and see what happens to her. So far, so good. I've been teaching her to get by on her brains and charisma.
LT et al: I've put in much larger samples of my writing, and I'm definitely male (by a relevant margin). This isn't voodoo; Gender Genie and the statistical analysis it's based on aren't perfect, but they're around 80% accurate.
WH: Oh, and how old is your daughter? :)
Perhaps it is a difference in the genre that the post is selected.
I took the last entry in HAB and put it through GG as 'blog entry' and it came out Male (4199/2125).
Same entry put through GG as 'non fiction' was completely different: Female 2455/1615.
Third time as 'fiction': Female 2455/1615
(as a control, I took the last entry from my blog, and put it through the same test, and it came out male for all genres (full disclosure, I'm male).
So perhaps the 'blog entry' for 'female' criteria is skewed to be something more 'Breakfast at Tiffany's' like than a philosophical discussion...
I'm just saying...
Gator: Interesting. The keywords don't lead me to believe they'd give context-sensitive results. Plus, you should use a llarger sample; I used the whole front page of the blog.
As both fiction and non-fiction, the author is very slightly female; but as a blog entry, the author is very solidly male.
For myself, analyzing the front page of my blog shows me male in all three categories, but less so in fiction and non-fiction than in blog entry.
It looks like those two categories may not be as well-trained as the third.
If it is a training problem.. then the latter could be skewed by a plethora of 'and then me and my friend went and talked about guys and yadda yadda...'
I think there is a more distinct difference between a 'journally blog', and a political/philosophical/current events type blog...
given the subject matter, I'd say that using 'blog post' for her/his/its writing isn't accurate, as it reads almost too academic to be show horned into blogisms...
Given the sample of your writing, that is also of a higher standard (in my opinion) than most blog posts, that you scored male in the f/nf genres, while HAC scored female might be more telling...
or it just might be pure fantasy on my part.. :)
just another thought...
to use the 'blog post' genre, a better sampling of 'bloggy goodness' might be found by mining the comments of several of the posts, and then having the genie evaluate that...
I'm late to not work.
I can't believe anyone is still talking about the Gender Weenie. Mine has a higher accuracy rate, and consumes far less resources.
X: Well, they're using a simplified version of the algorithm. 71% is still pretty good, and I expect the actual strength of any given predition is dependent on how skewed the gender ratio is.
G: I imagine that the three categories are arbitary and simply used for training purposes. I doubt they function any differently, they probably just weight the words a bit differently.
Actually, that was my point. Did you try MY version of the Gender Weenie? Based on the comments, so far it has batted 1.000.
As an aside, even if 71% meant 71%, it still wouldn't be that impressive, when you consider that a 50% accuracy rate can be attained simply by flipping a coin.
X: Yes I tried it, very nice :)
And 71% is pretty good, much better than could be done over time with a random coin flip.
I'm not so sure I trust the thing--I'm a girl, and when it tried it on my own blog (ironically enough, a set of entries we're I'm actually babbling about female stuff, like wearing skirts in spring and being maid of honor for a wedding), it classified me as mail. There's not guarantee the modified version is even 71% accurate--and even if it is, there's still a 29% inaccuracy rate. That's hardly a guarantee that HAC's some guy.
SC: Yes well, I was joking about absolute certainty :) And I'm sorry to hear that you're merely a piece of mail.