Uncovering the Bias: Can We Recognize Gay Faces from Photos?

[zoobyshoe]

A study found that students asked to tell whether someone was gay or straight guessed correctly more often than could be put down to mere chance.
Women had greater accuracy with 65 per cent able to identify someone's sexuality at a glance, while men were correct 57 per cent of the time.
Evidence suggest it is easier to recognise gay women's faces than men's even when photos were shown upside down and with no hairstyle visible.
Researchers in journal PLoS One say the results suggest we may unconsciously make gay or straight decisions when meeting a new face.
Joshua Tabak, of the University of Washington, said: "It may be similar to how we don't have to think about whether someone is a man or a woman or black or white.

http://www.telegraph.co.uk/science/science-news/9269298/Women-really-do-have-a-gaydar.html

My question: what photos were shown, and do they demonstrate a confirmation bias in selecting only masculine looking women as the lesbian examples?

A couple weeks ago I was going through my pictures trying to put together an album of the lesbians I have photographed in preparation for a possible show of these pics at a coffeehouse in a neighborhood known for its larger gay population. I found I had to avoid selecting shots that happened to enhance more masculine aspects of them. That is, I started off unconsciously selecting the most masculine looking shots.

If I put together an album of those masculine looking shots and then added a bunch of random, very feminine looking shots of women I know not to be lesbians, I could easily stack the deck such that almost anyone would be able to pick out the lesbians. Is this what happened in this study?

My own "gaydar" is highly dependent on hearing how someone talks, how they dress and their hair, and how they move. It's only about 50% accurate, too, in that some people make a point of telegraphing their sexuality and others completely avoid it. Unless they "stacked the deck" as I suggested, I very much doubt it's possible to tell if someone's gay from the shape of their face alone and in a still photograph.

 

[QuarkCharmer]

I agree with you. Outside of the stereotypical nuances one associates, I don't think there is any possible way to determine the preference of a person. What was the sample size of this study? They said that the percent correct was way to high to be mere coincidence, but I wouldn't believe that unless I saw the numbers myself. I mean come on, what makes a face "gay"?

I was going through my pictures trying to put together an album of the lesbians I have photographed

Furthermore, I would like to see your photo album

 

[zoobyshoe]

I agree with you. Outside of the stereotypical nuances one associates, I don't think there is any possible way to determine the preference of a person. What was the sample size of this study?

You mean this? :

For the study, 129 college students viewed 96 photos each of young adult men and women who identified themselves as gay or straight.

Furthermore, I would like to see your photo album

Here's one I posted a couple years back. Just for fun, there's one gay guy and one lesbian in this group. You pick 'em out.

https://www.physicsforums.com/showthread.php?t=306587

 

[Ryan_m_b]

Seems like this study was very poorly done. Firstly why just two sexualities? Why not use one of the many sexuality scales like the Kinsey scale? Whilst there is an argument to be put forth that there are fashions in queer culture that are distinct it is a pretty week one considering all the diversity of fashion inherent within society. The fact that gaydar doesn't exist is monumentally unsurprising.

 

[zoobyshoe]

The fact that gaydar doesn't exist is monumentally unsurprising.

Gaydar exists. About 50% of the time I can tell if someone is gay without them directly telling me. This is because they deliberately 'telegraph' the fact, because, I think, they want to be recognizable to other gay people.

Some people do not pay attention and never make the connection between deliberately telegraphed indicators and another's sexuality. They have no "gaydar".

Some people, though, simply do not 'telegraph' any information about their sexuality.

 

[SHISHKABOB]

gaydar is like radar

a dude pings out a signal, and he interprets that signal

I don't think it would work too well with just a picture of a face. But if you sit down and chit chat with someone for a while, then it *may* become apparent, even if it's not explicit. But ofc, I don't think I could figure if a person from another culture was gay or not, because it's all very much determined by the social norms that I'm aware of.

 

[lisab]

Gaydar exists. About 50% of the time I can tell if someone is gay without them directly telling me. This is because they deliberately 'telegraph' the fact, because, I think, they want to be recognizable to other gay people.

Some people do not pay attention and never make the connection between deliberately telegraphed indicators and another's sexuality. They have no "gaydar".

Some people, though, simply do not 'telegraph' any information about their sexuality.

When I was (very) young, I knew I had excellent gaydar.

Then I learned not every gay person fits the stereotype.

Then I further learned, *most* gays don't fit the stereotype.

Now I know my gaydar sucks.

 

[Ryan_m_b]

When I was (very) young, I knew I had excellent gaydar.

Then I learned not every gay person fits the stereotype.

Then I further learned, *most* gays don't fit the stereotype.

Now I know my gaydar sucks.

Very true. And conversely not every heterosexual person fits a stereotype. Metrosexuality as it's called it's pervasive in many fashion groups.

 

[zoobyshoe]

When I was (very) young, I knew I had excellent gaydar.

Then I learned not every gay person fits the stereotype.

Then I further learned, *most* gays don't fit the stereotype.

Now I know my gaydar sucks.

I know what you mean, but I would say it's straightdar that most often is unreliable. The assumption that, in the absence of gay signals someone must be straight is where most people err. Some people just don't telegraph anything about their sexual preference, and the assumption they are all straight is the one most likely to be wrong.

 

[mcknia07]

I'm pretty good at this! I can spot one out in a minute, well... ok some are harder than others, but most of them just have that vibe.

 

[mrdebraj]

Leaving apart the people who fit into / adopt the stereotype, its not very reliable to predict someone's sexuality.

Also its true that some continuous variation between gay and straight exists (someone mentioned the Kinsey scale). Often the subject oneself is confused or even unaware of ones' own fluctuations...

I myself am always missed out of other people's dars :p

 

[Evo]

Who new Jillian Michaels is gay?

http://tv.yahoo.com/news/-biggest-loser--trainer-jillian-michaels-becomes-a-mother-of-two.html

 

[Borg]

Who new Jillian Michaels is gay?

http://tv.yahoo.com/news/-biggest-loser--trainer-jillian-michaels-becomes-a-mother-of-two.html

Or http://news.yahoo.com/jim-parsons-comes-gay-230559700.html? My wife claims to have good gaydar but she was surprised by that.

 

[zoobyshoe]

Also its true that some continuous variation between gay and straight exists (someone mentioned the Kinsey scale). Often the subject oneself is confused or even unaware of ones' own fluctuations...

People consciously misreport themselves a lot, too, probably for simplicity's sake. To be accurate most people who characterize themselves as "gay" probably should be saying "bi".

In any event, the notion you can tell anything definite from a photograph strikes me as hard to believe.

 

[zoobyshoe]

Or http://news.yahoo.com/jim-parsons-comes-gay-230559700.html?

He set my gaydar off in about 2 seconds. Seriously.

edit: here's my reaction to him the first time Dave posted a clip of the show:

https://www.physicsforums.com/showpost.php?p=2491043&postcount=152

 

[256bits]

I do not think the conclusion of the study has any merit in its attempt to describe what the general population possesses in terms of gaydar.
Its conclusion should have been that a small young (?) university volunteer group of mostly female members was able to discern, from altered (clipped) facial photographs selected from Facebook by another group of unknown individuals ( whose demographics and preferences are unknown), the sexual preference of the individual depicted in said photograph correctly by a factor somewhat just greater than by chance.
And that further study on the general population should be done to confirm or deny the premise obtained.

Personnally I do not think the study adds anything more to the scientific pool of information since it was so limited.

As humans tend to label and categorize everything around them why should that not be surprising that gay/not gay is a category to try to fit someone else into. If someone has gaydar than that would mean they would have to rely on vague clues from previous experience and certainly that is not foolproof. Rumour had it at one time that John Wayne was gay.

 

[turbo]

My question: what photos were shown, and do they demonstrate a confirmation bias in selecting only masculine looking women as the lesbian examples?

When I got settled into college over 40 years ago, the hottest-looking woman on campus was the president of the Wilde-Stein Club (gay social organization) and her lover was arguably the second-hottest (though close). One day, I was killing some time in the student union lounge with my Gibson 12-string, playing and singing, and they both sat in the love-seat facing mine. They stayed quite a long time, and as my flamboyantly gay friend from my area of Maine entered the lounge, the ladies started hugging and sharing passionate kisses. My friend told me that I was "another victim" or something similar, suggesting an ambush. I had no idea. My friend told me who they were, and he was laughing at my expense. I didn't have any gaydar then.

I had no idea of their sexual orientation or club affiliation, and my friend was laughing his butt off.

 

[zoobyshoe]

I didn't have any gaydar then.

It's an acquired skill, for sure. Two of the coffee houses where I hang out are owned by gay men and also project a counter-culture image, which attracts a large gay clientele to them. I happened to meet one lesbian girl who was an art student in one of them. For a year and a half, she regarded me with suspicion, but finally sat down and drew with me one day. We became best pals, in fact, and I ended up meeting all her lesbian pals. Other lesbians who saw me sitting with lesbians started approaching me cold. Now I know billions of lesbians. Without making any conscious effort to do so, I started picking up on tell-tales, and now I believe my lesbian gaydar is better than most.

As I said earlier, I think people fail when they think they have straightdar: the ability to tell someone is straight or "not-gay". That picture Evo posted looks like a straight woman to me. A lot of lesbians are "Lipstick Lesbians". They present as really pretty, girly, straight women. I can't tell a Lipstick Lesbian from a straight girl at all: they telegraph no clues whatever that I can perceive.

 

[turbo]

I'm not out much in public anymore (even in open-air places), but when I was playing in local taverns, there was a lesbian who loved my custom Harley. She had a sport-bike, and every time she showed up for shows, she had the cutest huggers on the back... She was a little on the stocky side (not fat) and had short hair, so we were on the same wave-length pretty quickly. She used to try to buy me drinks, even though the bar-maids kept me well-supplied. Sweet lady. I don't know how she smuggled those huggers into the taverns - most looked under-aged.

 

[Illuminerdi]

It's an acquired skill, for sure. Two of the coffee houses where I hang out are owned by gay men and also project a counter-culture image, which attracts a large gay clientele to them. I happened to meet one lesbian girl who was an art student in one of them. For a year and a half, she regarded me with suspicion, but finally sat down and drew with me one day. We became best pals, in fact, and I ended up meeting all her lesbian pals. Other lesbians who saw me sitting with lesbians started approaching me cold. Now I know billions of lesbians. Without making any conscious effort to do so, I started picking up on tell-tales, and now I believe my lesbian gaydar is better than most.

As I said earlier, I think people fail when they think they have straightdar: the ability to tell someone is straight or "not-gay". That picture Evo posted looks like a straight woman to me. A lot of lesbians are "Lipstick Lesbians". They present as really pretty, girly, straight women. I can't tell a Lipstick Lesbian from a straight girl at all: they telegraph no clues whatever that I can perceive.

I have excellent straightdar. It's over 95% accurate!

 

[Evo]

I would have guessed that Justin Bieber was gay.

 

[zoobyshoe]

I have excellent straightdar. It's over 95% accurate!

Can you tell from photographs?

 

[zoobyshoe]

I would have guessed that Justin Bieber was gay.

Same here. He looks like most of the lesbians I know.

 

[Illuminerdi]

Can you tell from photographs?

I think you're missing the joke.
If you say "straight" every time, you'll be right an overwhelming majority of the time.

 

[phylotree]

It is hard, and perhaps unnecessary to prove whether one is gay or not. I have met people (men) of 30-50 years old, married and having children (lovely boys and girls), but gay is gay. They think it's a shame and they might probably lose several things altogether if that is known. If Justin Bieber were gay, news corps would earn more money.

 

[zoobyshoe]

I think you're missing the joke.
If you say "straight" every time, you'll be right an overwhelming majority of the time.

I would have got it, but 95% is too high a success rate to quote. Most people don't think 95% of people are straight.

http://www.gallup.com/poll/6961/what-percentage-population-gay.aspx

My perception is more like 20% are gay.

 

[Illuminerdi]

I would have got it, but 95% is too high a success rate to quote. Most people don't think 95% of people are straight.

http://www.gallup.com/poll/6961/what-percentage-population-gay.aspx

My perception is more like 20% are gay.

I've heard stats that say 10% before, but I'm not actually seeing them in real life, and I live in Chicago, not the South. I'm wondering if "gay" is being used generally. I've never heard 20%, though.

 

[Ken Natton]

It seems to me that whatever the accuracies or scientific rigour of this study, what it shows is that, at best, ‘gaydar’ is an extremely blunt instrument. And I would suggest there is one highly accurate and vary reliable indicator of sexual orientation. If someone tells you that they are homosexual, then the overwhelming probability is that they are homosexual.

If the question is of weighing up the possibility of forming a loving relationship with someone you don’t know well enough to discuss sexual orientation with, then sexual orientation is only one factor in the equation, and people have long had methods of sending out signals and establishing mutual interest before laying their cards on the table. Yes, some are better at that than others. Some unfortunate people repeatedly lay their cards on the table before the wrong people. But again, sexual orientation is not the only source of their mistakes.

And yes, despite modern greater acceptance of people’s freedom to express their sexuality how they may, still some choose to play their cards close to their chest. But it seems to me that the spirit of the game is to keep an open mind about such people, and to close your mind on the basis of an unwarranted faith in the abilities of your ‘gaydar’ is as foolish as any other form of prejudice.

 

[zoobyshoe]

I've heard stats that say 10% before, but I'm not actually seeing them in real life, and I live in Chicago, not the South. I'm wondering if "gay" is being used generally. I've never heard 20%, though.

I don't know Chicago at all. Is there a part of town considered to be the "gay" section? Here in San Diego the neighborhood, Hillcrest, is widely known as being the part of the city mostly inhabited by gay people. It's kind of a "Little San Francisco". If there were such a neighborhood in Chicago then figuring the percentage of Chicago's gay population would entail getting a sense of how large the population of that section is. Hillcrest, I have become aware, has burst it's borders and now, in all the neighboring areas you see a lot of same sex couples walking down the street holding hands. I've seen the same thing happening in Golden Hill, which is far flung from Hillcrest, but near a college and which has a lot of businesses that cater to that age group. Younger people are less afraid to be in the open about it. My guesstimate is 20%.

 

[Illuminerdi]

I don't know Chicago at all. Is there a part of town considered to be the "gay" section? Here in San Diego the neighborhood, Hillcrest, is widely known as being the part of the city mostly inhabited by gay people. It's kind of a "Little San Francisco". If there were such a neighborhood in Chicago then figuring the percentage of Chicago's gay population would entail getting a sense of how large the population of that section is. Hillcrest, I have become aware, has burst it's borders and now, in all the neighboring areas you see a lot of same sex couples walking down the street holding hands. I've seen the same thing happening in Golden Hill, which is far flung from Hillcrest, but near a college and which has a lot of businesses that cater to that age group. Younger people are less afraid to be in the open about it. My guesstimate is 20%.

I'm not sure what the official name of the neighborhood is, but everyone refers to it as "boy's town", but I don't think the numbers for Chicago are anywhere near 20%, when accounting for it.

 

[zoobyshoe]

I'm not sure what the official name of the neighborhood is, but everyone refers to it as "boy's town", but I don't think the numbers for Chicago are anywhere near 20%, when accounting for it.

I've been assuming San Diego is "normal" in it's percentage, but it may not be. It could be it is attracting a larger gay population than most cities, becoming more like San Francisco. I don't know. I see openly gay couples where ever I go here but I only go where the coffee houses are.