The Gender Genie Blinks

Ok, this is just for fun: Inspired by an algorithm developed by Moshe Koppel, Bar-Ilan University in Israel, and Shlomo Argamon, Illinois Institute of Technology, the Gender Genie attempts to predict the gender of the author.

Tip: Test it with 500 words of text or more in order to get best results. When I tried it with a few hundred words, it guessed wrong, but it guessed right with more than 500+ words.

Apparently the algorithm gives a various-weighted round of points to FEMININE keywords (With, If, Not, Where, Be, When, Your, Her, We, Should, She, And, Me, Myself, Hers, and Was) vs. a various-weighted scale to MASCULINE keywords (Around, What, More, Are, As, Who, Below, Is, These, The, A, At, It, Many, Said, Above and To).

One of the conclusions I was able to determine is that this algorithm assumes men write with a style that is *informational* and women write with a style of *involvement*. What do you think?

18 Comments »


1
Zack writes:

Hi Christopher M. Knight,

This is a interesting article :) First time knowing that they are able to know the gender of the author by using maths.

I will agree with you that men write with a style that is *informational* I do not know about other men but it does apply to me.

When I write my article, I will prefer a style that will be more informational so that the reader will get the information immediately.

Nice post of this information as it is interesting :)

Comment provided May 17, 2007 at 2:58 PM

[Reply]

2

It’s wonderful! I am decidedly male in my writing! Here are my results: Decidedly male! Text from an informational page on how to write a novel, so it has a presentation format to convince people of value so as to buy product.

Am I male from it, or is the marketing style I use to convince and sell masculine?

They take key common words like and and total them to indicate the female and male scores.
————-
Words: 1222

Female Score: 1192
Male Score: 1879
————-new test text————-
Text from a sharing email to a woman friend:
Words: 1419

Female Score: 2221
Male Score: 2174
The Gender Genie thinks the author of this passage is: female!
—————
It worked again with closer scores, but still! The point is that I am in my feminine side in sharing personally with a female friend and it shows up in the style of writing I do which involves mainly personal sharing and understanding of the other person.

This sort of proves the Jungian psychology point that the inner dynamic is the archetype of masculine and feminine being in both sexes. If teenagers naturally identify with their genital and cultural development stereotypes they will show themselves to either go with their genitals or go to the opposite in psychological development and behavior.

What this means is that the archetypes, the inner masculine and feminine, are somewhat independent of genitals and hormones. This little test indicates that a ‘real study’ can be done to prove the presence of masculine and feminine differences regardless of sex, but maybe tied or not tied to identity.

In my case since I am a man developing both masculine and feminine aspects of myself I then am able to ‘turn on’ the feminine in a feminine endeavor, that of either trying to convince, which is masculine, or opening to relate, share, connect, which is feminine.

Am I able to show both sides because of the difference in the kind of writing, or because of the ability to express both sides of my inner selves, despite having only male genitals?

This little tool may be quite effective in delimitating inner and even outer dynamics and styles.

I don’t put a lot of work here into being scientific, but at the fun level it was sort of amazing.

Maybe they can devise a similar test that will indicate if a person is high or low on prosperity consciousness?

For business it might help writers to do the first part of an article in the relational feminine and then with relationship to reader established go into a ‘hard sell’ decisive masculine. And you test this ability to write both styles out of yourself and in writing copy.

Successful copy may be one or the other. I shall take more room here and test my one article that far outstrips other of my articles.

Comment provided May 17, 2007 at 3:26 PM

[Reply]

3

Utterly fascinating!!!

Results on my article about naked massage in the Gender Genie is as follows:
————
Words: 1399

Female Score: 2899
Male Score: 2195
The Gender Genie thinks the author of this passage is: female!
————
My writing style is decidedly feminine, according to the genie, in this article and I have over 6000 views for this article and only 3445 views for all my other sixteen articles.

Does this suggest that writing in a feminine style, whether male or female, will make you a much more popular writer to be read?

The implications are absolutely staggering!

All that is left to do is test my other articles, starting with the least read and see if they tend towards the masculine. This I won’t do just now. Back to my other writing!

Thanks for this, Chris!!!

Comment provided May 17, 2007 at 3:39 PM

[Reply]

4

Strephon,

With a day of programming, we probably could add to your article reports if the tone is female or male, but I’m not certain the data would provide any actionable value beyond entertainment.

I’m also pretty sure that the topic of the content is not relevant in this algorithm as much as the words that are used to glue each sentence together or the reason for the choice in words (relation vs. information).

Comment provided May 17, 2007 at 3:51 PM

[Reply]

5

This is rather fascinating. The first two articles I tried definitely came out female. Then I tried a couple longer articles and both came out male. I think I’m having a bit of an identity crisis here!

Strephon makes an interesting point: Does the male or female voice affect readership?

Thanks for the interesting distraction this afternoon.

Comment provided May 17, 2007 at 4:43 PM

[Reply]

6
Jan writes:

Words: 1479

(NOTE: The genie works best on texts of more than 500 words.)

Female Score: 1580
Male Score: 1769

The Gender Genie thinks the author of this passage is: male!

This piece is part of an adventure novel I’m currently writing for teenaged boys (Hardy Boys style)

~~~~~~~~~~~

Words: 882

(NOTE: The genie works best on texts of more than 500 words.)

Female Score: 1582
Male Score: 857

The Gender Genie thinks the author of this passage is: female!

(This is a non-fiction article written for adult Christian women.)

~~~~~~~~~~~~~~

Words: 648

(NOTE: The genie works best on texts of more than 500 words.)

Female Score: 1129
Male Score: 900

The Gender Genie thinks the author of this passage is: female!

(The Introduction portion of an ebook – about how to make money writing.)

~~~~~~~~~~~~~~

Words: 804

(NOTE: The genie works best on texts of more than 500 words.)

Female Score: 977
Male Score: 956

The Gender Genie thinks the author of this passage is: female!

(another fiction piece, for young/middle adult years – probably women)

Perhaps, the determination of which gender is who you’re writing for?)

Jan

Comment provided May 17, 2007 at 5:02 PM

[Reply]

7
Sharon writes:

Looks like I’m a man – I tested it on three blog posts. Who knew?

Comment provided May 17, 2007 at 6:42 PM

[Reply]

8
Lance Winslow writes:

I have another set of words for REAL MEN:

Crush Annihilate smash win action compete attack score rush pulverize mangle destroy kick punch

And the man authors should be able to reach high places and pick-up heavy objects. Additionally the man author should be able to think of articles to write while envisioning how to find something while driving without asking for directions!

Who needs artificial intelligence author gender recognition when you have a Lance around to jab, joust, stick, poke, and thrust at your adversary, competition, enemy, opponent or fellow man?

Comment provided May 17, 2007 at 9:14 PM

[Reply]

9
Jan writes:

Wooohooo! Lance, did you take your testosterone shot today, or what?

If you’ll take off that pink shirt and those hot pink stillettos you can stop singing soprano too!

Jan

Comment provided May 17, 2007 at 9:17 PM

[Reply]

10
Lance Winslow writes:

No actually this is my normal self but consider I am writing an eBook on:

“Military Unmanned Ground Vehicles Tactics”

this week, I have using my male vocabulary to the fullest. You know I use to have a four page four column set of “action” words I use to read to myself before track meet competitions. I think it helped me win, but it may have been just lots of training, genetics and mental focus? I would read it once before leaving the house it psyches you up to win. Fun stuff.

My question would be about the gender AI scheme that it might change by culture too. But each type language would have its own words and each class of each culture their own slang. I wonder if you could make one of these things for Instant Messaging or Text Messaging. Or one to find violent behavior to catch terrorists? I bet someone has, so may be I should be careful what I write on the blog? Hmmm?

Comment provided May 17, 2007 at 9:25 PM

[Reply]

11

Lance, you sweet, little slaver! I love you!!!!

Comment provided May 18, 2007 at 3:31 AM

[Reply]

12

Christopher,

Actually I am quite serious about adding scientific analysis to writing. You have an enormous resource here for research as to what good writing is, and how you define that of course. You are in control of an excellent stable of good writers whose minds are well developed in information processing.

However, you have competition, and will have more of it and more powerful competition. Top companies and corporations do research. Google does research on its methods and data production.

Right now you are an information directory but you don’t know your impact, your effectiveness, your value as a publisher of information on the web. This is what I get from you, anyway. Yet you were trained in psychology, you said before.

You don’t seem to do research. You don’t seem to know what you are doing with your numbers. You don’t seem to know what your numbers mean.

I suggest the other side of the coin is that this Genie is a serious tool. Scientists made the algorithm. They don’t do this for fun. I am trained as a scientist. We use these tools and statistics to accurately define effects from causes.

I am quite serious in wanting to know why one of my articles gets over 6000 views while all the 17 others only get 3500 views about.

I am quite serious when I want scientific evidence on writing effectiveness.

I want to know if Lance, the writer of the largest number of articles here, is also the writer who has the best average of reader-per-article, or which sections that he writes for get the highest reader-per-article response. He has a great mind. What is his effectiveness also? How can we measure this? For all us writers, we want to write better and better.

If you can program in software to get us some real statistics for analysis by scientifically trained people you are helping the effectiveness of your writers and the reputation of your company.

Google is scientific. You can be also, but you have to have a scientific research team analyzing this tremendous resource of writers working free for you. They would love to know back what their effectiveness is.

And so would you! Right?

More effectiveness equates greater impact on the web and in the culture. Right?

These are not critical comments but probing comments about the values of doing scientific research to better a company such as yours. Research and Development, it is called. Re-search what you’ve got.

Comment provided May 18, 2007 at 4:51 AM

[Reply]

13

Strephon,

Since you are a scientist :), have you done any experiments to determine what the key factors might be that determine writing effectiveness?

Comment provided May 18, 2007 at 7:38 AM

[Reply]

14

Chris,

This is what excites me here with the little Genie and the whole burst of the web which I will not live to see that long.

In working on the scientific end of what is writing effectiveness I had training in liguistics, and what they discovered is remarkable. But I had to work in the real world, so I chose psychology as my profession. I learned how inaccurate psychological testing can be. I learned how scientifically you cannot prove that psychotherapy works.

I learned how difficult it is to do objective research. But in applying scientific method to dreamwork and people’s lives changing I was left only with clinical reporting of results which is in my books.

Not a great much has happened yet in scientific studies of dreaming and dreamwork. Not much at all.

It’s as if the government money all goes to the military for research and development, and we can see how much more effective in killing people they do become.

So when I see Google devoting its best brains to researching how to best display information for people searching, I am in joy that maybe entreprenuers like yourself will use some of your money, if you earn it, to become scientific in trying to objectify results.

My learning as a writer has been learning techniques, what works best in writing novels that sell. However, the industry as a whole is very unknowligiable. When seven out of ten books published fail to make money, this is a terrible state. In auto manufacturing all cars are built to work.

So I say, and it was in a recent New York Times article, that in writing we don’t know what makes writing effective, so says the New York Times. They don’t know even today what makes for a best seller.

And also they say that publishers do not do research. Marketing people do because in their business you have to use methods that work to get people to buy your products. But if you are only publishing ideas, and don’t have to get people to buy them, like here maybe, then you don’t have to do research.

This is where I am coming from. I sell a 92,000 words collection of over 2000 writing tools and techniques. I have brought the collection together from many writing craft books and my own discoveries. I sell it to help novelists, but I cannot say at this point how much it will help. If however I start getting feedback and money for my efforts I will certainly report what seems to work best for writers and why.

That’s called improving the product. I will use scientific research to discover how to use these. Now my biggest measure of writing success is sales: 350,000 books in circulation. That is so crude, but it is real world. I hope for other measuring devices that work so I can be more effective.

Hope this helps.

Comment provided May 18, 2007 at 8:19 AM

[Reply]

15

Strephon,

We compile incredible digital reports on all kinds of stats designed to help us measure and manage by metrics.

Whenever we find a metric that has meaning and can provide actionable assistance to our members, we selfishly have an incentive to provide it because we both win when your articles perform better.

The thing I’m pretty certain about is that if TRAFFIC to an article is the measure of success that we’re discussing, that there will always be variables outside of your and our control that may significantly influence those stats without any reporting or our/your knowledge of the origin of the traffic.

Example: An outside party likes one of your articles and blogs about it. You get an instant 2k page views spike.

Example: An outside email newsletter publisher promotes your article to his or her 50,000 member ezine causing either your article or your website to get a spike of traffic.

Example: A visitor of our site likes one of your articles and lists it in his Stumbleupon account. 400 people who visit this users Stumbleupon account surf your article because they trust Stumbleupon to help them discover content that they most likely will be interested in based on their profile or history.

I could go on… :)

Comment provided May 18, 2007 at 8:55 AM

[Reply]

16

Last night this post led me to think about being authoritative in one’s articles.

Is “Authoritative” more of Male (dominant, objective, dealing with facts) than Female (passive, subjective, dealing with feelings)?

Next I was thinking of how one might become more authoritative in one’s articles on purpose (mine AND my clients’ articles, that is). Is it by writing with a more “male” tone?

I was also pondering, is the art of persuasion male, or female? At first I though it must be female. But then I was thinking that if your argument is based in emotion, it’s female. If it’s based in fact, it’s male.

To me, fact weighs in heavier than emotion, but emotion drives most purchasing decisions.

So I now believe that the “tone” of advertising is a feminine or subjective one, but the “tone” of journalism is a “male” or objective one. And I do believe you can take an authoritative stance no matter which style of writing you choose.

Does this make ANY sense?

Comment provided May 18, 2007 at 10:29 AM

[Reply]

17
Lance Winslow writes:

A very wise person once told me that the total difference between all men and all women was less than the range of differences between all women. And therefore I would hesitate to categorize anhy woman you did not know and even then would consider it unwise to categorize any woman or all women using some silly computer algorithm, because humans although somewhat predictable are not so easily figured out and if someone ever does figure out all women, please email me.

Comment provided May 20, 2007 at 10:42 PM

[Reply]

18
Susan Scharfman writes:

Aurore Dupin, the lover of Frederic Chopin, wrote under the pseudonym of George Sand. That was the 18th century. Ha! The end of the 20th century saw J.K. Rowling doing the same thing. Her target audience is boys, so she was advised by her first publisher (Bloomsbury) to use only first initials in her name or she wouldn’t make a shilling. That proved right in Britain after the first printing. But it didn’t matter at all when Scholastic bought Harry Potter. Does Rowling write male or female? Neither. Nora Roberts is the richest most prolific author in her romance genre. When she writes under her Sci-fi name of J.D. Robb, she tends to use the colder metallic words needed for that genre, yet she is female.

My point is: I don’t think it matters whether you’re writing is inclined to be male or female, if there is such a thing. Certainly you should be able to switch back and forth for a novel. I think successful novel writers have their fingers on the pulse of what sells, and churn it out for mass market. When writing an informative article or writing to teach what others can use, I tend to use more authoritative words, strong verbs and quotes, than when I am sharing an experience.

Is there a mathematical solution to every problem? I’m lousy at it. I prefer Wayne Dyer’s.

Comment provided May 24, 2007 at 12:52 PM

[Reply]

RSS feed for comments on this post.

Leave a comment

Please read our comment policy before commenting.