# Turing test for women?

**URL:** <https://boards.straightdope.com/t/turing-test-for-women/793719>\
**Category:** In My Humble Opinion\
**Created:** [August 15, 2017, 4:06pm UTC](https://boards.straightdope.com/t/turing-test-for-women/793719 "2017-08-15T16:06:21Z")\
**Posts on this page:** 20\
**Page:** 2

<div class="post-metadata">

**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\
**Post date:** [August 16, 2017, 1:55am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/21 "2017-08-16T01:55:31Z")

</div>

> [@Hari\_Seldon](#):
>
> …  
> Do you think you could tell a man from a woman that you communicate with only by computer? Do you think most people could? One thing I do know is that unless a Doper makes it clear by choice of UID or by the personal nature of his/her posts, I do not know the sex of many Dopers.

Back to the OP and away from the red herring of job interviews …

Given enough time and a wide enough range of natural topics I’d get a bunch, but by no means all, correct. As I think most people would.

OTOH, until \*\*Infovore \*\*outed herself up-thread I thought she was male. We’ve had little to no direct back-and-forth interaction here, but I’d read enough of her words on various topics that she’d moved from my personal Doper default of “person, presumably male” to “male person”. Oops on me. :o

I agree that one of the things I like about SDMB is that for many folks their sex is so submerged that I can process them as a person, not as a man or a woman. It was a refreshing, and somewhat unsettling, wakeup call back when I was a newbie to find out that so-and-so was male or female when I had earlier decided or (worse?) assumed, the opposite.

Which really raises the much larger question of how different society would be if everybody really was just a person and their male/femaleness was no more important to interactions than is their eye color or ear shape?

---

<div class="post-metadata">

**Author:** ![Robot\_Arm](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/robot_arm/32/18280_2.png) [@Robot\_Arm](https://boards.straightdope.com/u/Robot_Arm)\
**Post date:** [August 16, 2017, 2:04am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/22 "2017-08-16T02:04:53Z")

</div>

> [@msmith537](#):
>
> Isn’t such a test “gender biased” by definition?

I wasn’t quite sure how to phrase that. People seem to be focusing on finding a question that would be more likely to be answered correctly by women than men, or vice versa. I’m wondering the opposite, are there questions that are answered by men and women at exactly the same correct percentage? If we could figure out what things men and women exactly agree on, it might be a clue to how a Turing test could tell them apart.

If that makes any sense.

---

<div class="post-metadata">

**Author:** ![DPRK](https://avatars.discourse-cdn.com/v4/letter/d/4491bb/32.png) [@DPRK](https://boards.straightdope.com/u/DPRK)\
**Post date:** [August 16, 2017, 3:45am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/23 "2017-08-16T03:45:26Z")

</div>

Couldn’t you deliberately defeat any such test by deliberately altering your writing style?

In response to **Hari Seldon** , I do not think you could very reliably tell a person’s gender from a Turing-style remote interview, but, especially if you are a robot, you could certainly make statistical guesses.

---

<div class="post-metadata">

**Author:** ![Walton\_Firm](https://avatars.discourse-cdn.com/v4/letter/w/45deac/32.png) [@Walton\_Firm](https://boards.straightdope.com/u/Walton_Firm)\
**Post date:** [August 16, 2017, 3:46am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/24 "2017-08-16T03:46:03Z")

</div>

> [@LSLGuy](#):
>
> Back to the OP and away from the red herring of job interviews …

The OP had two paragraphs; the first one was about job interviews. You quoted his (I think **Hari Seldon** is male, though now that I think about it, I’m not sure what I base that on) second paragraph, replacing the first one with “…”, and then declared job interviews to be a red herring distracting us from the OP?

---

<div class="post-metadata">

**Author:** ![Nava](https://avatars.discourse-cdn.com/v4/letter/n/da6949/32.png) [@Nava](https://boards.straightdope.com/u/Nava)\
**Post date:** [August 16, 2017, 4:25am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/25 "2017-08-16T04:25:35Z")

</div>

> [@Infovore](#):
>
> Most people who’ve only communicated with me via computer (including in World of Warcraft) think I’m a guy unless I tell them. I don’t make any effort to mislead them–I just tend to have a fairly direct style of writing.

One of my favorite horrid remarks:  
“You can’t be a girl, you’re too good a tank!”

OK, there’s no girlz in tha interwebz and apparently typing on a keyboard requires upper body strength to be done correctly.

---

<div class="post-metadata">

**Author:** ![Nava](https://avatars.discourse-cdn.com/v4/letter/n/da6949/32.png) [@Nava](https://boards.straightdope.com/u/Nava)\
**Post date:** [August 16, 2017, 4:36am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/26 "2017-08-16T04:36:33Z")

</div>

> [@Robot\_Arm](#):
>
> I wonder if you could come up with a question that _wouldn’t_ have a gender bias in who was most likely to answer correctly.

Anything where the answer is a piece of factual data. Not everybody may remember that Columbus discovered America on October 12, 1492 (and one may nitpick the “discovered” part) or that the atomic number of oxygen is 8 and its atomic weight 16 (and one may complain that it’s actually 15.99something, with the something changing depending on calculation methods), but they are still questions with factual responses that have an educational bias but not a gender bias.

---

<div class="post-metadata">

**Author:** ![asterion](https://avatars.discourse-cdn.com/v4/letter/a/4da419/32.png) [@asterion](https://boards.straightdope.com/u/asterion)\
**Post date:** [August 16, 2017, 4:38am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/27 "2017-08-16T04:38:59Z")

</div>

> [@Robot\_Arm](#):
>
> I wonder if you could come up with a question that _wouldn’t_ have a gender bias in who was most likely to answer correctly. It’s like that logic puzzle where you’re on an island where half the people lie and half tell the truth.
> 
> How many home runs did Babe Ruth hit, divided by the number of times Liz Taylor was married?

Without looking anything up, is that 714/6?

---

<div class="post-metadata">

**Author:** ![Infovore](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/infovore/32/420_2.png) [@Infovore](https://boards.straightdope.com/u/Infovore)\
**Post date:** [August 16, 2017, 4:44am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/28 "2017-08-16T04:44:06Z")

</div>

> [@Nava](#):
>
> One of my favorite horrid remarks:  
> “You can’t be a girl, you’re too good a tank!”
> 
> OK, there’s no girlz in tha interwebz and apparently typing on a keyboard requires upper body strength to be done correctly.

Haha! Since my two mains (demon hunter and death knight) are both tanks, that made me laugh. 🙂

---

<div class="post-metadata">

**Author:** ![Trafalgar\_Laura](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/trafalgar_laura/32/2907_2.png) [@Trafalgar\_Laura](https://boards.straightdope.com/u/Trafalgar_Laura)\
**Post date:** [August 16, 2017, 5:29am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/29 "2017-08-16T05:29:53Z")

</div>

> [@DPRK](#):
>
> Couldn’t you deliberately defeat any such test by deliberately altering your writing style?

Although I doubt how really accurate such tests are, I have a lot of fun with the tests on the Internet that claim to take samples of your writing and tell you your gender. I can generally game them, and I imagine others could too. Then again, I write fiction (more as an avocation than a vocation) so I just think of writing in a ‘male’ voice rather than a ‘female’ voice. It’s hard for me to define why I feel they’re different, but at least these websites are reassuring me that I’m doing all right at it…

Back to the OP, whatever the actual results were (more or less minority hiring,) it does seem like some blind way to measure skills and history would be a rather useful way to hire. Still, it would remove the interview stage where one can decide if a person is really a fit for the job and community, or would be a problem or irresponsible. I think this personal contact can be very important too… but it does open the door to conscious or subconscious bias leading to the decision that a woman or black person “just isn’t right.”

ETA: Looking back, I’m not sure the hiring issue really was the point of the OP entirely, but on the whole I think my post is germane…

---

<div class="post-metadata">

**Author:** ![Colibri](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/colibri/32/1841_2.png) [@Colibri](https://boards.straightdope.com/u/Colibri)\
**Post date:** [August 16, 2017, 5:33am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/30 "2017-08-16T05:33:05Z")

</div>

> [@UDS](#):
>
> Yes, we know. The point is that in most job interviews that will not be a relevant or proper question to ask.
> 
> The questions need to be directed that the requirements/specifications for the post. Unless knowledge of the character of Carrie Bradshaw and her footwear preferences is relevant to the job requirements, the question seems to be put for the purpose of eliciting the gender of the interviewee (based on stereotypes about topics that interest women but not men) and the interviewee can properly object to the question (and, when not offered the job, bring an action alleging employment discrimination on the grounds of sex).

As has been pointed out, the OP mentions job interviews, but the actual question is about distinguishing men from women purely on the responses to computer questions.

As I said, certain questions can give you clues, even though they don’t provide anything like certainty. Asking about the number of pairs of shoes you own is probably one of them. Asking about the number of power tools you own is probably another.

[Some studies have identified differences in the writing styles used by men and women.](http://articles.chicagotribune.com/2003-05-28/features/0305280247_1_interactive-style-writing-anonymous-author)

---

<div class="post-metadata">

**Author:** ![Walton\_Firm](https://avatars.discourse-cdn.com/v4/letter/w/45deac/32.png) [@Walton\_Firm](https://boards.straightdope.com/u/Walton_Firm)\
**Post date:** [August 16, 2017, 6:00am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/31 "2017-08-16T06:00:46Z")

</div>

> [@Trafalgar\_Laura](#):
>
> Although I doubt how really accurate such tests are, I have a lot of fun with the tests on the Internet that claim to take samples of your writing and tell you your gender.

[Here](http://www.hackerfactor.com/GenderGuesser.php) is one such test, in case people want to play with it. Apparently, a simple algorithmic text analysis does better than chance, but not very much better: the author claims it is 60-70% accurate. Also, apparently European men tend to get classified as “weakly male” by the computer.

---

<div class="post-metadata">

**Author:** ![Robot\_Arm](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/robot_arm/32/18280_2.png) [@Robot\_Arm](https://boards.straightdope.com/u/Robot_Arm)\
**Post date:** [August 16, 2017, 6:01am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/32 "2017-08-16T06:01:06Z")

</div>

> [@Nava](#):
>
> Anything where the answer is a piece of factual data. Not everybody may remember that Columbus discovered America on October 12, 1492 (and one may nitpick the “discovered” part) or that the atomic number of oxygen is 8 and its atomic weight 16 (and one may complain that it’s actually 15.99something, with the something changing depending on calculation methods), but they are still questions with factual responses that have an educational bias but not a gender bias.

I think even lots of factual questions would turn out to have a bias in the answers. Do you think equal percentages of women and men would know the atomic number of oxygen?

Note that the difference may well be due to educational or social bias of some sort, but that’s beside the point that I’m trying to make. What questions could help you tell a man from a woman, and are there any questions that would really be no help at all.

> [@asterion](#):
>
> Without looking anything up, is that 714/6?

No. I’ll give you a hint, the answer is not an integer.

---

<div class="post-metadata">

**Author:** ![Nava](https://avatars.discourse-cdn.com/v4/letter/n/da6949/32.png) [@Nava](https://boards.straightdope.com/u/Nava)\
**Post date:** [August 16, 2017, 6:33am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/33 "2017-08-16T06:33:01Z")

</div>

> [@Robot\_Arm](#):
>
> I think even lots of factual questions would turn out to have a bias in the answers. Do you think equal percentages of women and men would know the atomic number of oxygen?
> 
> Note that the difference may well be due to educational or social bias of some sort, but that’s beside the point that I’m trying to make. What questions could help you tell a man from a woman, and are there any questions that would really be no help at all.

Differences if any are _completely_ educational. Ask students at my college c. 1975, and correct responses would have come completely from males - because there were no women, it’s the year the first female student was admitted. Ask people from my own class in 6th, 10th, 11th, 12th grade or in college and you’ll get equal percentages. And I don’t mean ask us when we were in those classes, I mean ask us now, years after we took those classes.

Factual answers tell you whether someone learned something; they do not tell you that someone’s gender.

---

<div class="post-metadata">

**Author:** ![Trafalgar\_Laura](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/trafalgar_laura/32/2907_2.png) [@Trafalgar\_Laura](https://boards.straightdope.com/u/Trafalgar_Laura)\
**Post date:** [August 16, 2017, 6:33am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/34 "2017-08-16T06:33:51Z")

</div>

Questions like “How many shoes do you own?” do lend some weight to a person being female… but I only have three pairs of shoes (and pretty much only wear one unless I have to dress up more.) And plenty of men are shoe-hogs. Similarly, some women like power tools, and my father explicitly has asked people to stop giving him tools under the assumption he wants them because he’s a guy. These also seem more like cultural cliches; are men in Uzbekistan known for loving power tools and grilling outdoors? (That’s a random country as an example, not a genuine request for knowledge about that specific place.)

I don’t mean these things wouldn’t be useful to a degree, just that as a ‘turing’ test it wouldn’t be much proof, and predictions based on questions that lean on stereotypes will have a decently high failure rate too.

And, as other ladies in this thread have noted, I often get mistaken for a male when I post under a gender-neutral username, particularly in more male-dominated environments like gaming.

---

<div class="post-metadata">

**Author:** ![Trafalgar\_Laura](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/trafalgar_laura/32/2907_2.png) [@Trafalgar\_Laura](https://boards.straightdope.com/u/Trafalgar_Laura)\
**Post date:** [August 16, 2017, 6:51am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/35 "2017-08-16T06:51:24Z")

</div>

This site had some interesting information about women in schooling, and what knowledge might be useful to determine gender based on education: [Current State of STEM | National Girls Collaborative Project](https://ngcproject.org/statistics)

If more girls take advanced science classes during high school than males, given the relatively introductory level of knowing the atomic number of common elements like oxygen, it seems as if that question would be a sign suggesting a female rather than a male.

That said, given the higher rate of women having degrees in social or biological sciences and of men having degrees in engineering or computer science, these might be a little more helpful if asking questions trying to determine gender by likely educational experiences. (I want to add the caveat that I’m strictly speaking in terms of education received, not about the potential learning ability of women in any field.)

Still not perfect… I do know several women who work in programming or otherwise in computer science.

---

<div class="post-metadata">

**Author:** ![eburacum45](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/eburacum45/32/8690_2.png) [@eburacum45](https://boards.straightdope.com/u/eburacum45)\
**Post date:** [August 16, 2017, 8:26am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/36 "2017-08-16T08:26:43Z")

</div>

Anyone who knows a little about Turing’s life would know that his so-called ‘Turing Test’ was originally based on a parlour game called _The Imitation Game_, which was intended to be a test to see if men and women could be distinguished on the basis of written answers.

[QUOTE=Chronos]  
Similarly, is the interviewee attempting to make es sex known, attempting to conceal es sex, or making no effort one way or the other?  
[/quote]

I use gender neutral pronouns a lot. For the possessive case I use ‘eir’, which is derived from ‘their’ - a gender neutral use of the plural to signify a singular pronoun which is often used in spoken English. ‘es’ seem unnatural to me, while another alternative, ‘er’, seems slightly too feminine for an epicene pronoun.

---

<div class="post-metadata">

**Author:** ![Nava](https://avatars.discourse-cdn.com/v4/letter/n/da6949/32.png) [@Nava](https://boards.straightdope.com/u/Nava)\
**Post date:** [August 16, 2017, 10:04am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/37 "2017-08-16T10:04:34Z")

</div>

> [@Trafalgar\_Laura](#):
>
> This site had some interesting information about women in schooling, and what knowledge might be useful to determine gender based on education: [Current State of STEM | National Girls Collaborative Project](https://ngcproject.org/statistics)
> 
> If more girls take advanced science classes during high school than males, given the relatively introductory level of knowing the atomic number of common elements like oxygen, it seems as if that question would be a sign suggesting a female rather than a male.

OK, here my own educational bias is showing. The atomic number of oxygen is required national curriculum in Spain in 6th grade (the last year of elementary school).

For people of my age and on the college track, it was required national curriculum in 6th and 10th; optional in 11th and 12th (you had to be science track to have chemistry) but the gender ratio in “humanities” and “sciences” tracks was 50:50, you needed to drill further down to find gender imbalances.

I specifically chose something which, while sciency, is basic enough to be something that most college-oriented people and many who weren’t aiming for college but who got compulsory schooling will have had to study at some point.

---

<div class="post-metadata">

**Author:** ![Bricker](https://avatars.discourse-cdn.com/v4/letter/b/977dab/32.png) [@Bricker](https://boards.straightdope.com/u/Bricker)\
**Post date:** [August 16, 2017, 11:19am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/38 "2017-08-16T11:19:04Z")

</div>

> [@Nava](#):
>
> Who is Carrie Bradshaw and I’m sorry but what’s that got to do with implementing SAP?

“How would you use custom messages DUX or DUY to create customer-specific events, and what brand of shoes would most distract you during the process?”

😃

---

<div class="post-metadata">

**Author:** ![Nava](https://avatars.discourse-cdn.com/v4/letter/n/da6949/32.png) [@Nava](https://boards.straightdope.com/u/Nava)\
**Post date:** [August 16, 2017, 11:26am UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/39 "2017-08-16T11:26:30Z")

</div>

I’m Ops, not Sales, and the brand of shoes most likely to distract me is any on my feet at the time. Does this question mean we’ll be allowed WFH? I like being able to work barefoot, but people in offices tend to frown on it.

---

<div class="post-metadata">

**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\
**Post date:** [August 16, 2017, 1:23pm UTC](https://boards.straightdope.com/t/turing-test-for-women/793719/40 "2017-08-16T13:23:32Z")

</div>

> [@Nava](#):
>
> OK, here my own educational bias is showing. The atomic number of oxygen is required national curriculum in Spain in 6th grade (the last year of elementary school).  
> …

Not disagreeing with any of your well-made points on sex/gender bias.

Ref the snip above, IMO your \*cultural \*bias is showing.

Here in good 'ol 'Merica, the corresponding standard is that 6th graders must have learned there is such a thing as “air”. In about 1/2 the states they also need to know that Jesus created it fairly recently. 😃

[Previous page](https://boards.straightdope.com/t/turing-test-for-women/793719.md?page=1)

[Next page](https://boards.straightdope.com/t/turing-test-for-women/793719.md?page=3)
