Male blogger, female blogger
Have you ever heard of the Gender Genie? If you’re a writer (as all bloggers are!), this is a fun exercise. The Gender Genie analyses the words you use and deduces whether you are male or female.
I experimented with my post on quadriforms because the Gender Genie prefers passages of 500 words or more. Click on the image to see a full size version of the results.
Here is the bottom line:
Words: 769
Female Score: 494
Male Score: 1292
The Gender Genie thinks the author of this passage is: male!
And the Gender Genie — just give me a second here —
[undoes pants, takes a look]
And the Gender Genie is correct! [much whistling, applause, general laughter]
OK, it’s a roundabout way of determining your sex. But the interesting part is the language analysis. Key words are isolated and identified as masculine or feminine. Each key word is assigned a value. For example, “with” is a feminine word and each use is valued at 52 points; “to” is a masculine word and each use is valued at only 2 points.
In some cases, I can follow the Gender Genie’s reasoning. “Your”, “her”, “we” and “myself” are all feminine words. “The” and “a” are both masculine words. The assumption is, women tend to discuss interpersonal matters relatively often; men discuss objects more often.
Surprisingly, “was” is feminine while “is” is masculine. Women use the past tense relatively often and men the present tense? Who knew?!
Does it work? Obviously the Gender Genie correctly identified my sex. For a woman’s result, check out MaryP’s blog.
The Gender Genie « It’s Not All Mary Poppins said,
Sunday, April 29, 2007 at 10:32 am
[...] Gender Genie Stephen found this site which will analyze a piece of writing and tell you, based on the frequency and [...]
juggling mother said,
Sunday, April 29, 2007 at 1:13 pm
Cool idea.
Sadly i am male:-)
only by a few words though, but since I really am quite a typical female in most of my outlook/abilities etc, it shows the basic principle is flawed.
Carolyn said,
Monday, April 30, 2007 at 12:03 pm
I’m curious to run one of my posts through that. Many people who come to my home have said it’s decorated gender ambiguous, and I feel I’m generally considered more verbally assertive than many of my gentler sexed counterparts (translation – I guess I can be a little bitchy…although if I were a man I think it would be looked at differently
).
Carolyn said,
Monday, April 30, 2007 at 12:49 pm
Yeah, it thought I was male. That’s ok…I like to keep em guessing.
Stephen said,
Monday, April 30, 2007 at 2:34 pm
I wonder if it ever mistakes a man for a woman? So far, it guesses a lot of women wrong, including MaryP about 50% of the time.
Ozymandias said,
Tuesday, May 1, 2007 at 12:15 pm
Interesting. I entered six different posts. It correctly identified me as a male 4 out of the six times.
Bill said,
Wednesday, May 2, 2007 at 12:12 am
Stephen you may find this interesting. I tried the Gender Genie on my stuff and came up Male every time (not a surprise) but then for the fun of it I ran some of George Eliot’s (pen name of Mary Ann Evans) poetry through it and what do you think it said?
Mary Ann Evans not only took on a male pen name but according to the Gender Genie she writes like a man. That said I thought poetry may not work in the genie so I tried poetry by Charlotte Bronte and she comes up female.
Handy little tool this. It makes you wonder if There was more to the pen name George Eliot than just a name?
Bill said,
Wednesday, May 2, 2007 at 12:17 am
Oh and I should mention I did it several times with Evans and Bronte just to be sure.
Bill said,
Wednesday, May 2, 2007 at 12:33 am
I tested the remaining Bronte and Evans poems and they both come up 50/50.
However I tested several other male poets and most came up 60 to 70 % Male except Edgar Allen Poe who is 50/50 or more female
This thing seems to be flawed on its ability to capture female writers.
Bill said,
Wednesday, May 2, 2007 at 12:38 am
The Oxford Journal of literary and linguistic computing says the following on automated text categorization. It would seem that this model does not live up to the technology which they claim is 80% accurate.
Automatically Categorizing Written Texts by Author Gender
Moshe Koppel1, Shlomo Argamon1,2 and Anat Rachel Shimoni1
1 Department of Mathematics and Computer Science, Bar-Ilan University, Israel 2 Department of Computer Science, Jerusalem College of Technology, Israel
The problem of automatically determining the gender of a document’s author would appear to be a more subtle problem than those of categorization by topic or authorship attribution. Nevertheless, it is shown that automated text categorization techniques can exploit combinations of simple lexical and syntactic features to infer the gender of the author of an unseen formal written document with approximately 80 per cent accuracy. The same techniques can be used to determine if a document is fiction or non-fiction with approximately 98 per cent accuracy.
Stephen said,
Wednesday, May 2, 2007 at 9:53 am
This thing seems to be flawed on its ability to capture female writers.
That’s the conclusion I’m coming to. It’s not to be taken very seriously, in my opinion, but it provides some passing amusement.
Jamie said,
Wednesday, May 2, 2007 at 10:54 pm
I tried it on about 8 different samples of my own writing, and it thinks I’m male too (on all but two samples). I’m with the rest of you guys: It doesn’t seem very accurate for female writers. But it was a good way to kill 20 minutes (not that I had 20 minutes to kill…!).
Bill, you seem to have invested quite a bit of time in the genie trying out all those poets!
Not telling said,
Monday, May 7, 2007 at 10:16 am
I’m trans, and gender genie guessed my gender correct 80% of the time. I’ve tested it on several friends (guys), and their test results were conclusive. I’ve noticed that this particular algorithm becomes particularly inaccurate when we used samples that were less than 400 – 600 words. Not only that, but technical non-fiction writing will be inaccurate at less than 2000 words.
Female writers are hard to pin down, as our writing style can change with topic, mood, and intended audience Women who work in a male dominated field will, inevitably, write like a guy. My mother’s work papers turned up male, while her personal writing ended up mostly female.
Here’s an interesting test for yous. I found this while rummaging through BBC.
http://www.bbc.co.uk/science/humanbody/sex/add_user.shtml
Remember. Gender is what’s between your ears, not what’s between your legs.
Stephen said,
Saturday, May 12, 2007 at 9:11 am
• Not telling:
I certainly take your point about writing style changing depending on context.
And I agree with your remark about gender, too. I admire C.G. Jung’s construct of the anima and the animus, and likewise the Taoist idea that we need to keep the male and female parts of our psyches in balance.
Not telling said,
Wednesday, May 16, 2007 at 10:02 am
Carl Jung was a very smart man. His theories appeal to me more than Freud’s. Freud’s theories seem to be based around paraphillias. In the end, he was just diagnosing himself, forcing his diagnosis upon everyone else as a generalization. It’s a shame and wonder that anyone would still listen to him. Oh well. Can’t be helped, I guess.
Writing style can change depending on age. I noticed as I was helping my mother grade papers, you could easily tell who was male or female. 12 to 14 year olds paper’s were the worst. Girls’ papers were very disjointed, didn’t stay on topic within paragraphs, and would go off on random tangents; guys’ papers were very strait to the point. Both were equally stupid. On occasion, I would get my grubby little hands on a paper that was written by someone smart. At that point, I would have to look towards handwriting style, or, god forbid, their name. Handwriting doesn’t have too much weight, as I know plenty of women who have horrendous handwriting.
In this case, I’m writing more objectively, so this post will probably be male according to gender genie… yep. Just goes to show, you know?