Yes, I do think it's fascinating if we can tell complex psychological information about an individual just by looking at them, the mechanisms underlying such determinations (if they exist) would seem to be fundamental questions about human perceptions. If you don't want to discuss them because they aren't interesting to you, fair enough - don't.

Discussing the validity of an experiment determining the truth of whether people do possess such a faculty seems like a perfectly normal thing to do, and being interested in it has little to nothing to do with whether some individual "prefers dicks or fannies".

Thommo wrote:Yes, I do think it's fascinating if we can tell complex psychological information about an individual just by looking at them, the mechanisms underlying such determinations (if they exist) would seem to be fundamental questions about human perceptions. If you don't want to discuss them because they aren't interesting to you, fair enough - don't.

Discussing the validity of an experiment determining the truth of whether people do possess such a faculty seems like a perfectly normal thing to do, and being interested in it has little to nothing to do with whether some individual "prefers dicks or fannies".

Fair enough.

Thommo wrote:Yes, I do think it's fascinating if we can tell complex psychological information about an individual just by looking at them, the mechanisms underlying such determinations (if they exist) would seem to be fundamental questions about human perceptions. If you don't want to discuss them because they aren't interesting to you, fair enough - don't.

Discussing the validity of an experiment determining the truth of whether people do possess such a faculty seems like a perfectly normal thing to do, and being interested in it has little to nothing to do with whether some individual "prefers dicks or fannies".

Yeah I agree. The importance of this research is not only in the existence of some kind of "gaydar" (which is interesting but probably not overly profound), but in the entire field of perception and psychology in general. Surely the introduction of the topic presented in the paper is enough for everyone to see how relevant this information is?

Not to mention the fact that when people bemoan boring or useless scientific studies, it's usually because (not being in the field themselves) they can't see the implications of the work. There was a study a while back that was reported in the media as developing an equation on the appropriate number of crisps a packet can contain before it reaches some significant level of breakage. I think it was reported in the Daily Mail or something so readers were outraged, as they routinely are, but it later turned out that this simple equation not only resulted in billions of dollars being saved worldwide by these crisp companies, but also had substantial implications for environmental issues related to crisp manufacture and packaging (something like maximising the number of crisps per package limited the amount of waste, which had huge implications when applied on a global scale).

Sure, it might be more interesting if all scientific results were about rockets and explosions, but sometimes the research with the most important effects on the world is that with the seemingly mundane or irrelevant findings.

FWIW I wasn't attacking the research, although I find its claim that people can tell whether someone else is gay with a success rate "greater than chance" rather underwhelming.

I suppose I was struck by the word "gaydar" which is often used in the less scientific sense of trying to work out whether people we know are gay - and particularly whether all sorts of other spurious associations (their speech, the music they like etc.) should confirm or deny such a conclusion. Outside of sexual pursuit, I just don't see why people bother, and TBH I can't help feeling there's something judgmental behind it all.

To use an analogy from a form of discrimination that's had slightly more history of being challenged - suppose we went to a party, and met someone who had just the slightest hint of negro features in their face, skin colouring etc. We then swapped phone numbers with this person and arranged to meet again. After they left, I said to you "hmm... Do you think that guy's mixed race?"

You'd probably think "why the fuck are you bringing that up apropos of nothing, and why do you care?" And if I was still talking about it three hours later, trying to work out whether the features that struck me really show he's mixed race, or whether I've also seen them in perfectly white people I know, and insisting that it's really important I know this as part of my ongoing relationship with the guy. Well then, we know what most people would think about my elevating such a thing to the status of "important", even if I don't do so in overtly judgmental terms.

I'm not saying every straight person who talks about the gaydar is a homophobe. I just think there's a residual discriminatory element of needing to put gay people in a controllable category and view our relationships with them through that category, in the fact of caring about it in the first place.

Beatsong wrote:I just think there's a residual discriminatory element of needing to put gay people in a controllable category and view our relationships with them through that category, in the fact of caring about it in the first place.

I am reminded of an essay by the late Isac Asimov from the 60s - remember this was in an era when long hair on men was an affront - where he discussed the common complaint, "You can't tell the boys from the girls" and continued by asking, Why is it necessary to tell the gender of a person at a glance? He was, as far as I remember, of the opinion that the reason was indeed discriminatory.

Beatsong wrote:FWIW I wasn't attacking the research, although I find its claim that people can tell whether someone else is gay with a success rate "greater than chance" rather underwhelming.

I don't think it's underwhelming, given that they were able to make these accurate predictions after only viewing the photograph for a fraction of a second (slightly above what is classed as subliminal). In real life we'd expect the accuracy to be far higher, given that you get more than a split second to look at someone, they are in colour (as opposed to the black and white photos from the study), and you get more cues (like voice and body language). So the fact that they were still scoring around 60% accuracy seems fairly impressive.

Beatsong wrote:I suppose I was struck by the word "gaydar" which is often used in the less scientific sense of trying to work out whether people we know are gay - and particularly whether all sorts of other spurious associations (their speech, the music they like etc.) should confirm or deny such a conclusion. Outside of sexual pursuit, I just don't see why people bother, and TBH I can't help feeling there's something judgmental behind it all.

I assume you know that the term "gaydar" wasn't used in the scientific article, so I agree that it carries with it a number of connotations that probably made it a bit inappropriate to use.

Beatsong wrote:To use an analogy from a form of discrimination that's had slightly more history of being challenged - suppose we went to a party, and met someone who had just the slightest hint of negro features in their face, skin colouring etc. We then swapped phone numbers with this person and arranged to meet again. After they left, I said to you "hmm... Do you think that guy's mixed race?"

Well, maybe it's not a good time for me to point out that studies on identification of things like race and mixed-race status are also performed. Generally they don't focus on whether they can identify races or not (since "race" is a physical characteristic so it's not surprising that people can identify physical characteristics), but more about the difference in success between different cultures and their ability to identify race, how different perceptions of race affect judgement of attractiveness, etc.

But the point is that research is done as well. This "gaydar" research is more interesting though given that sexual orientation is not a physical characteristic, so it's pretty amazing that we can be so accurate in identifying it by just looking at a face for a fraction of a second.

Beatsong wrote:You'd probably think "why the fuck are you bringing that up apropos of nothing, and why do you care?" And if I was still talking about it three hours later, trying to work out whether the features that struck me really show he's mixed race, or whether I've also seen them in perfectly white people I know, and insisting that it's really important I know this as part of my ongoing relationship with the guy. Well then, we know what most people would think about my elevating such a thing to the status of "important", even if I don't do so in overtly judgmental terms.

I'm not saying every straight person who talks about the gaydar is a homophobe. I just think there's a residual discriminatory element of needing to put gay people in a controllable category and view our relationships with them through that category, in the fact of caring about it in the first place.

I guess so. It's certainly a problem when people make an issue of it, as if being able to identify a gay person was nearly a matter of life or death, etc, but the fact that the ability does exist makes it an interesting thing that will inevitably come up in discussion (and certainly relevant of scientific investigation). I suppose it's similar to the argument that there are obvious differences between men and women - pointing out that difference and discussing it is not a problem, but it becomes a problem when it becomes justification for treating them differently, especially in areas that are not directly relevant to the supposed differences.

I suppose that's sort of what you're saying, right? That investigating the existence of this ability isn't a problem, but the problem is with how this "gaydar" is applied in real life and the importance that's attached to it?

Mr.Samsa wrote:

Beatsong wrote:FWIW I wasn't attacking the research, although I find its claim that people can tell whether someone else is gay with a success rate "greater than chance" rather underwhelming.

I don't think it's underwhelming, given that they were able to make these accurate predictions after only viewing the photograph for a fraction of a second (slightly above what is classed as subliminal). In real life we'd expect the accuracy to be far higher, given that you get more than a split second to look at someone, they are in colour (as opposed to the black and white photos from the study), and you get more cues (like voice and body language). So the fact that they were still scoring around 60% accuracy seems fairly impressive.

Did you read my criticism of the first experiment a couple of pages back? The accuracy was not 60% as far as I can tell, there appears to be some subtle errors in the statistical analysis. What do you think?

Yeah, I haven't had a chance to read through it properly yet though. I just meant at face value the result didn't seem insignificant or anything to scoff at - but obviously that's just assuming the result was true.

Mr.Samsa wrote:Yeah, I haven't had a chance to read through it properly yet though. I just meant at face value the result didn't seem insignificant or anything to scoff at - but obviously that's just assuming the result was true.

Yeah, we are in agreement on that point! I am slightly sceptical though, I really can't see a mechanism here. Replication of the result would be most interesting.

Mr.Samsa wrote:I suppose that's sort of what you're saying, right? That investigating the existence of this ability isn't a problem, but the problem is with how this "gaydar" is applied in real life and the importance that's attached to it?

Pretty much, yeah. I have no problem with them conducting the research. Depending on how the results are interpreted it could be viewed as a study into what kinds of things lead viewers to assume someone is gay, as much as a study of what kinds of characteristics are gay. It's all a perfectly valid area of enquiry.

I just feel wierd when I am involved in or overhear long winded conversations about "whether so-and-so is gay or not". It's a bit like I was saying on the thread about gender a while back that I think you contributed to: I have trouble with the idea that one aspect of a person's psycho-sexual makeup (in this case, whether they are most likely to want to have sex with men or women) necessarily brings with it a whole load of other stuff that puts them in a larger "category". And I have particular trouble with the idea that it's important to know what that category (which was largely made up in the first place) is, as part of getting to know the person.

Just to be clear I'm not having a go at the OP or anyone else. It's just a personal statement of something that mystifies me about people. But then there are a lot of things about people that mystify me.

Mr.Samsa wrote:Yeah I agree. The importance of this research is not only in the existence of some kind of "gaydar" (which is interesting but probably not overly profound), but in the entire field of perception and psychology in general. Surely the introduction of the topic presented in the paper is enough for everyone to see how relevant this information is?

Not to mention the fact that when people bemoan boring or useless scientific studies, it's usually because (not being in the field themselves) they can't see the implications of the work. There was a study a while back that was reported in the media as developing an equation on the appropriate number of crisps a packet can contain before it reaches some significant level of breakage. I think it was reported in the Daily Mail or something so readers were outraged, as they routinely are, but it later turned out that this simple equation not only resulted in billions of dollars being saved worldwide by these crisp companies, but also had substantial implications for environmental issues related to crisp manufacture and packaging (something like maximising the number of crisps per package limited the amount of waste, which had huge implications when applied on a global scale).

Since I brought it up in this thread, I'm assuming that you're directing this at me. You're arguing against a straw man. Researchers should continue to develop the body of scientific knowledge. However, the public does not have to concern itself with such topics unless they find it interesting and gain enjoyment from it. Since the current results are so limited, it isn't a social issue yet. That will change if and when the ability to perceive sexual orientation has been proven to be true and practically significant.

Mr.Samsa wrote:I don't think it's underwhelming, given that they were able to make these accurate predictions after only viewing the photograph for a fraction of a second (slightly above what is classed as subliminal). In real life we'd expect the accuracy to be far higher, given that you get more than a split second to look at someone, they are in colour (as opposed to the black and white photos from the study), and you get more cues (like voice and body language). So the fact that they were still scoring around 60% accuracy seems fairly impressive.

That's a terrible accuracy rate and practically unusable. For there to be practical use, rather than simply statistical significance, I would be looking for at least 80% (arbitraily chosen, simply as it is represents a vast majority of successes). Not to mention, certain people will have greater intuition (e.g. ability to read body language) than others, so averaging it across people isn't very helpful. Also, how do you separate the ability to read body language in general and the ability to sense sexual orientation?

Mr.Samsa wrote:But the point is that research is done as well. This "gaydar" research is more interesting though given that sexual orientation is not a physical characteristic, so it's pretty amazing that we can be so accurate in identifying it by just looking at a face for a fraction of a second.

It's impressive, but not unheard of for people to intuitively sense things in a split second, including for things that aren't physical.

Mr.Samsa wrote:I suppose that's sort of what you're saying, right? That investigating the existence of this ability isn't a problem, but the problem is with how this "gaydar" is applied in real life and the importance that's attached to it?

That's a very good point. In politically backward places with poor human rights records (e.g. Saudi Arabia), the findings could potentially be deadly, whether it is proven or not, so long as the authorities in question believe they are.

SeriousCat wrote:However, the public does not have to concern itself with such topics unless they find it interesting and gain enjoyment from it.

I don't see the point in this as a criticism it applies to quite literally everything. It seems relatively clear that there is no obligation on the general public to discuss topics that they find boring and unpleasant to discuss.

I also don't think it matches what you said before*, though if it's a revised or clarified position, fair enough.

*"Before engaging in any intellectual queries, that is drilling down the different orders of logic of different positions on a topic, one should ask themselves whether the topic is worth discussing."
"(1) How was this assertion verified; and (2) If the assertion was correct, what would be its significance. The first question you ask should be the one that is easiest to answer, thus potentially saving you from a lot of irrelevant work."

Thommo wrote:

Mr.Samsa wrote:Yeah, I haven't had a chance to read through it properly yet though. I just meant at face value the result didn't seem insignificant or anything to scoff at - but obviously that's just assuming the result was true.

Yeah, we are in agreement on that point! I am slightly sceptical though, I really can't see a mechanism here. Replication of the result would be most interesting.

Skeptical of being able to discriminate between sexual orientation using facial characteristics in general or being able to do so in the timeframe suggested in this study? The former is backed by a fairly large research base, so I assume the latter?

Beatsong wrote:

Mr.Samsa wrote:I suppose that's sort of what you're saying, right? That investigating the existence of this ability isn't a problem, but the problem is with how this "gaydar" is applied in real life and the importance that's attached to it?

Pretty much, yeah. I have no problem with them conducting the research. Depending on how the results are interpreted it could be viewed as a study into what kinds of things lead viewers to assume someone is gay, as much as a study of what kinds of characteristics are gay. It's all a perfectly valid area of enquiry.

Yeah that's what the research was looking at. In the article, they actually break down all possible ways that certain characteristics could act as cues for sexual orientation and attempted to figure out which characteristics were leading to the positive score.

SeriousCat wrote:

Mr.Samsa wrote:Yeah I agree. The importance of this research is not only in the existence of some kind of "gaydar" (which is interesting but probably not overly profound), but in the entire field of perception and psychology in general. Surely the introduction of the topic presented in the paper is enough for everyone to see how relevant this information is?

Not to mention the fact that when people bemoan boring or useless scientific studies, it's usually because (not being in the field themselves) they can't see the implications of the work. There was a study a while back that was reported in the media as developing an equation on the appropriate number of crisps a packet can contain before it reaches some significant level of breakage. I think it was reported in the Daily Mail or something so readers were outraged, as they routinely are, but it later turned out that this simple equation not only resulted in billions of dollars being saved worldwide by these crisp companies, but also had substantial implications for environmental issues related to crisp manufacture and packaging (something like maximising the number of crisps per package limited the amount of waste, which had huge implications when applied on a global scale).

Since I brought it up in this thread, I'm assuming that you're directing this at me. You're arguing against a straw man. Researchers should continue to develop the body of scientific knowledge. However, the public does not have to concern itself with such topics unless they find it interesting and gain enjoyment from it. Since the current results are so limited, it isn't a social issue yet. That will change if and when the ability to perceive sexual orientation has been proven to be true and practically significant.

Well I think Beatsong brought it up first, but either way it wasn't directed at anyone in this thread - it was a general comment sparked by the discussion between you and Thommo. With that said, I'm not sure why you think the public shouldn't be interested in such issues - they don't need to have practical, relevant, everyday social implications in order to be considered interesting. We already know that we have an ability to perceive sexual orientation, that isn't what this study demonstrated. This study attempted to extend on the knowledge base of previous research which demonstrated that we could perceive sexual orientation, and argued that we can make an accurate judgment within a near-subliminal period of time.

SeriousCat wrote:

Mr.Samsa wrote:I don't think it's underwhelming, given that they were able to make these accurate predictions after only viewing the photograph for a fraction of a second (slightly above what is classed as subliminal). In real life we'd expect the accuracy to be far higher, given that you get more than a split second to look at someone, they are in colour (as opposed to the black and white photos from the study), and you get more cues (like voice and body language). So the fact that they were still scoring around 60% accuracy seems fairly impressive.

That's a terrible accuracy rate and practically unusable. For there to be practical use, rather than simply statistical significance, I would be looking for at least 80% (arbitraily chosen, simply as it is represents a vast majority of successes). Not to mention, certain people will have greater intuition (e.g. ability to read body language) than others, so averaging it across people isn't very helpful. Also, how do you separate the ability to read body language in general and the ability to sense sexual orientation?

Again, the accuracy rating was for a near-subliminal amount of time looking at a black and white photograph. Any way you try to bend it, that's impressive.

The purpose of the study was to help determine what characteristics are used to determine sexual orientation, rather than simply asking whether we can distinguish between sexual orientations based on physical characteristics (because we already know that we can do that). Yes, individuals will differ, but I'm not sure what relevance that has. The same can be said of absolutely every between-subject study; for example, every study of medication will include individuals that will respond extremely well to the drug, and those that won't respond at all. Also, the study didn't look at body language at all (only black and white photographs), so here we've separated the ability to read body language from the ability to perceive sexual orientation.

SeriousCat wrote:

Mr.Samsa wrote:But the point is that research is done as well. This "gaydar" research is more interesting though given that sexual orientation is not a physical characteristic, so it's pretty amazing that we can be so accurate in identifying it by just looking at a face for a fraction of a second.

It's impressive, but not unheard of for people to intuitively sense things in a split second, including for things that aren't physical.

You reckon? I can't think of any study that has had such a high accuracy from using so few possible cues.

Mr.Samsa wrote:

Thommo wrote:

Mr.Samsa wrote:Yeah, I haven't had a chance to read through it properly yet though. I just meant at face value the result didn't seem insignificant or anything to scoff at - but obviously that's just assuming the result was true.

Yeah, we are in agreement on that point! I am slightly sceptical though, I really can't see a mechanism here. Replication of the result would be most interesting.

Skeptical of being able to discriminate between sexual orientation using facial characteristics in general or being able to do so in the timeframe suggested in this study? The former is backed by a fairly large research base, so I assume the latter?

In the timeframe and reduced information faces presented in this study specifically, yes.

Thommo wrote:In the timeframe and reduced information faces presented in this study specifically, yes.

Fair enough, I agree with that.

Looking back on your earlier criticism, I agree that it's certainly something that the authors should have included. Although not directly relevant to the question of being able to identify "homosexual faces", it would provide relevant information that could help determine what feature is being used to make such a judgement. On top of that, some other concerns I'd have would be:

1) they were using Facebook photos, which are often picked after careful consideration and posing, so could result in artificial design features that could cue people towards correctly identifying sexual orientation, and

2) all of the participants (the ones guessing the sexual orientation) were female. This doesn't really affect the details of the finding so much, but it would cause us to question this conclusion: "(b) that sexual orientation is inferred more easily from women’s vs. men’s faces". I think it's possible that women are better at identifying gay women, and men better at identifying gay men, rather than it necessarily being the case that gay women are more identifiable than gay men in general.

Apparently this was from a Master's student thesis work, so that probably explains some of the weird decisions they made in writing up their article. Presumably though, they should have the data on the other "hits", so you could probably email them and ask them for it.

I've just spent 30mins reading up on the techniques of signal detection theory and then 10mins with pen and paper checking the calculation of A' with either the H and F rates or the H' and F' rates and I was just wrong.

The H* rate isn't calculated how I thought it was - it's not the number of correct "gay face" guesses as a proportion of "gay face" guesses, it's the number of correct "gay face" guesses as a proportion of "gay faces", thus no data is disregarded and the answer would be the same whether you counted by "straight faces" instead (assuming that there were no faces which did not elicit a response, which I think was the case). Thus H* = 1-F and F* = 1-H. Subsitituing these in the equations for calculating A' results in the same equation in the co-ordinates of H* and F*.

So, mea culpa, the statistical treatment aspect is fine.

Mr.Samsa wrote:Looking back on your earlier criticism, I agree that it's certainly something that the authors should have included. Although not directly relevant to the question of being able to identify "homosexual faces", it would provide relevant information that could help determine what feature is being used to make such a judgement. On top of that, some other concerns I'd have would be:

1) they were using Facebook photos, which are often picked after careful consideration and posing, so could result in artificial design features that could cue people towards correctly identifying sexual orientation, and

2) all of the participants (the ones guessing the sexual orientation) were female. This doesn't really affect the details of the finding so much, but it would cause us to question this conclusion: "(b) that sexual orientation is inferred more easily from women’s vs. men’s faces". I think it's possible that women are better at identifying gay women, and men better at identifying gay men, rather than it necessarily being the case that gay women are more identifiable than gay men in general.

I think those are both highly relevant points though, thanks for the response.

Mr.Samsa wrote:Well I think Beatsong brought it up first, but either way it wasn't directed at anyone in this thread - it was a general comment sparked by the discussion between you and Thommo. With that said, I'm not sure why you think the public shouldn't be interested in such issues - they don't need to have practical, relevant, everyday social implications in order to be considered interesting.

I will say this again, as I have done so repeatedly: if you find it interesting and enjoyable then it becomes important to you. By all means enjoy it. Frontier science is often interesting. However, it does not become a social issue that is pressed for time and needs to be discussed in the public mediasphere. Right now it's just a collection of possibilities.

Mr.Samsa wrote:Again, the accuracy rating was for a near-subliminal amount of time looking at a black and white photograph. Any way you try to bend it, that's impressive.

I'm not trying to bend or distort anything, so that is clearly a straw man. You are assuming that the accuracy would dramatically increase to levels that would constitute practical significance with longer and closer examination of a persons face.

Mr.Samsa wrote:The purpose of the study was to help determine what characteristics are used to determine sexual orientation, rather than simply asking whether we can distinguish between sexual orientations based on physical characteristics (because we already know that we can do that). Yes, individuals will differ, but I'm not sure what relevance that has. The same can be said of absolutely every between-subject study; for example, every study of medication will include individuals that will respond extremely well to the drug, and those that won't respond at all. Also, the study didn't look at body language at all (only black and white photographs), so here we've separated the ability to read body language from the ability to perceive sexual orientation.

I didn't disagree with this, so I don't understand why you brought it up.

Mr.Samsa wrote:You reckon? I can't think of any study that has had such a high accuracy from using so few possible cues.

I'm not referring to studies of people in general, or inuition common across most people. Perhaps I shouldn't mention anecdotal evidence, but I can't help but remember people I know who are especially good at reading others. I'm sure most of us know of people who just have an uncanny ability to figure out what you're thinking with just a glance. Sure, it's not a scientific study, but it's a well known phenomenon. Even people without incredible intuitive talent can learn that intuition over time, like a old married couple knowing what the other person is thinking beyond mere behavioural patterns with just a glance.

SeriousCat wrote:

Mr.Samsa wrote:Well I think Beatsong brought it up first, but either way it wasn't directed at anyone in this thread - it was a general comment sparked by the discussion between you and Thommo. With that said, I'm not sure why you think the public shouldn't be interested in such issues - they don't need to have practical, relevant, everyday social implications in order to be considered interesting.

I will say this again, as I have done so repeatedly: if you find it interesting and enjoyable then it becomes important to you. By all means enjoy it. Frontier science is often interesting. However, it does not become a social issue that is pressed for time and needs to be discussed in the public mediasphere. Right now it's just a collection of possibilities.

And again, I'll point out that it needs to be discussed in public mediasphere because it's an interesting scientific result and people like to discuss science.

SeriousCat wrote:

Mr.Samsa wrote:Again, the accuracy rating was for a near-subliminal amount of time looking at a black and white photograph. Any way you try to bend it, that's impressive.

I'm not trying to bend or distort anything, so that is clearly a straw man.

It's a figure of speech, not an accusation.

SeriousCat wrote:You are assuming that the accuracy would dramatically increase to levels that would constitute practical significance with longer and closer examination of a persons face.

I'm not assuming anything, we already know that accuracy in identifying sexual orientation from facial features is much higher in experiments where the images are presented for longer than 40-50ms. Some of the research is even discussed in the paper we're currently discussing.

SeriousCat wrote:

Mr.Samsa wrote:The purpose of the study was to help determine what characteristics are used to determine sexual orientation, rather than simply asking whether we can distinguish between sexual orientations based on physical characteristics (because we already know that we can do that). Yes, individuals will differ, but I'm not sure what relevance that has. The same can be said of absolutely every between-subject study; for example, every study of medication will include individuals that will respond extremely well to the drug, and those that won't respond at all. Also, the study didn't look at body language at all (only black and white photographs), so here we've separated the ability to read body language from the ability to perceive sexual orientation.

I didn't disagree with this, so I don't understand why you brought it up.

Yes, you disagreed with all my points above. Firstly you claimed: "That's a terrible accuracy rate and practically unusable". This is one of the instances where you support the mistaken belief that this paper was trying to demonstrate that sexual orientation could be identified by looking at facial features. I corrected you by pointing out that we already know that this can be done, so attacking the "accuracy" rates of this in the real world is irrelevant, because obviously in the real world you'd get more information than just a black and white photograph, and a longer look than just 40-50ms. As such, if we were going to use an accuracy rating for the "real world", we'd use the much higher accuracies found in previous studies that aren't looking at subject responses in a split-second task.

Secondly, you claimed "Not to mention, certain people will have greater intuition (e.g. ability to read body language) than others, so averaging it across people isn't very helpful", and I corrected you by pointing out that even though large scale research designs can flatten out individuals differences, this isn't overly important (as demonstrated by the fact that nobody complains when individual differences are eliminated in medical trials).

Thirdly, although not technically a "disagreement", you asked: "Also, how do you separate the ability to read body language in general and the ability to sense sexual orientation?", and I pointed out that they successfully identified sexual orientation in a task that had no body language element at all.

Your response that you didn't disagree with anything I said, and that you were unsure as to why I said the things I did, is just perplexing.

SeriousCat wrote:

Mr.Samsa wrote:You reckon? I can't think of any study that has had such a high accuracy from using so few possible cues.

I'm not referring to studies of people in general, or inuition common across most people. Perhaps I shouldn't mention anecdotal evidence, but I can't help but remember people I know who are especially good at reading others. I'm sure most of us know of people who just have an uncanny ability to figure out what you're thinking with just a glance. Sure, it's not a scientific study, but it's a well known phenomenon. Even people without incredible intuitive talent can learn that intuition over time, like a old married couple knowing what the other person is thinking beyond mere behavioural patterns with just a glance.

I'd be highly skeptical of such anecdotal claims, given the amount of confounds and cognitive biases inherent within them. I'm also hugely skeptical that it's common in everyday life for people to only have 40-50ms access so a few cues and still be able to make accurate predictions of highly complex social phenomena.

I only skimmed the thread so I may have missed if someone else brought this up.

But... different groups of people have different fashions, different ways of presenting themselves, that are either purposefully or subconsciously identify them with various things. Use of makeup, facial expressions (poses, etc) and hairstyles are also subject to this ("this" meaning the peer influence of your friends/culture).

So my guess is that "gaydar" isn't inbuilt, it is more or less our ability to recognize cultural stereotypes. I would like to see this study done with the same data in Africa or Asia. The less westernized the better.

I have seen a snippet in Scientific American - Mind that mentioned that homosexual people can tell of other people's homosexuality through their smell, allowing people to find others who are attracted to each other. Sounds a bit far fetched and I didn't see any evidence or study behind it as it was a little snippet.

Has anyone else heard of this or have a arrow to point to some credible evidence supporting this claim? I'd love to hone it.