Do males and females exhibit different levels of variation in traits?

[Math Is Hard]

Chris Hillman mentioned something in the engineering forum recently that piqued my curiosity:
https://www.physicsforums.com/showthread.php?t=149937

Again, I stress that I don't neccessarily believe this as stated, but some cognitive scientists have suggested that, across species, males as a group tend to be more diverse that females as a group in various metrics, for fundamental reasons related to some subtleties in how natural selection operates. Indeed, they say, as group, men vary more in height, "general intelligence", whatever, than do women. In the sense of standard deviation. So that if they are right, even if the mean GRE score by males agrees very nearly with the mean score by females, if you look at students who scored very low on the GRE, and those who scored very high, in both cases you might see females under-represented.

Some questions came to mind:
Do males really vary more in traits than females? Which traits? Is this true for most species? What explanations are accepted for how this came about?

 

[Math Is Hard]

For the general intelligence aspect, what I've found related to variability and sex is a study done in 1932 in Scotland. It is discussed here in a debate between Steven Pinker and Elizabeth Spelke:
http://www.edge.org/3rd_culture/debate05/debate05_index.html

One other data set meeting the gold standard is displayed in this graph, showing the entire population of Scotland, who all took an intelligence test in a single year. The X axis represents IQ, where the mean is 100, and the Yaxis represents the proportion of men versus women. As you can see these are extremely orderly data. In the middle part of the range, females predominate; at both extremes, males slightly predominate. Needless to say, there is a large percentage of women at both ends of the scale — but there is also large sex difference.

Graph:http://www.edge.org/3rd_culture/debate05/images/pinker.slides/pages/pinker_Page_41.htm
So the guys get more geniuses, but also more idiots. Maybe... but to me, studies that old are a bit "suspect". However, the same phenomenon is apparently observed in SAT scores and another study that is mentioned in their debate.

But I don't want to dwell on measures of intelligence. I'm interested in other traits like height, and why males would vary more than females. Why would natural selection favor this?

 

[Evo]

I would ask Chris Hillman to provide the sources for his comments.

I have read some articles about human male - female variances in how the brain works, there does truly seem to be a difference there. Women are much better at multi tasking. Men are better at thinking "spatially" while women think better logically. I'm trying to find it.

 

[Math Is Hard]

I would ask Chris Hillman to provide the sources for his comments.

I did indeed. Looking forward to the response.

I have read some articles about human male - female variances in how the brain works, there does truly seem to be a difference there. Women are much better at multi tasking. Men are better at thinking "spatially" while women think better logically. I'm trying to find it.

I responded to that a little bit in the other thread. But what I am interested in here is the larger claim that males of a species are more "diverse" than females in certain traits - e.g., height, eye color, intelligence, what-have-you, and how the mechanisms of natural selection may have supported this.

I was thinking this might be relevant in the more sexually dimorphic species, but not in humans.

 

[Evo]

But what I am interested in here is the larger claim that males of a species are more "diverse" than females in certain traits - e.g., height, eye color, intelligence, what-have-you, and how the mechanisms of natural selection may have supported this.

I was thinking this might be relevant in the more sexually dimorphic species, but not in humans.

Yeah, that's a pretty broad statement and not one I have seen supported.

 

[Moonbear]

I would think it would be hard to demonstrate that's valid across a large number of traits in a large number of species. The amount of variation of a trait does typically reflect the amount of selection for that trait, but that doesn't necessarily mean more selection will result in only a "middle" range for that trait. Selection can also result in two extremes predominating without much in the middle.

Sexually dimorphic traits don't usually mean females are in the middle and males spread out over a large range, they usually mean females tend toward one extreme and males toward the other (i.e., on a scale of body sizes, you'll find more short women and more tall men, but in other species, the females may all be larger and the males all smaller). So, sexually dimorphic traits are clear examples where the means are not similar, and males are not exhibiting a greater range of traits than females, but instead, the two sexes are exhibiting distinctly different traits. This would indicate such a claim is not generalizable.

 

[Math Is Hard]

Thanks. :)

 

[LeonhardEuler]

I was just reading what you were saying and a possible hypothesis came to mind. Bear in mind I'm just guessing, but I know there are a lot of traits such as color blindness, which are rare in the general population but much more common in males than in females. This is because a male needs only one X-chromosome with a recesive trait in order to display that trait, whereas a female needs two. If there are genes controlling height, intelligence, etc. on the X-chromosome, one would expect males to exhibit more of the rare phenotypes. I don't know whether any genes controlling these things actually exist on the X-chromosome, so I'm really just guessing.

 

[Q_Goest]

Hey Math, great post by the way. I read this one last night and it's still buggin' me.

That graph you posted is a shocker. I read the description before opening it up. I fully expected to see two conventional bell curves, with the female curve just a bit taller and more narrow than the men's bell curve. But of course, the men's bell curve is upside down! I guess there's a wealth of genius men and a wealth of drooling morons as well, but few in between. I wonder if any more recent studies have confirmed this 'upside down' bell curve. It seems like one of those things you scratch your head over and say, "That just can't be!" I'd love to know if any more recent studies have been found to support the original one.

~

On a side note, have you noticed that almost all the winners of the Darwin Awards are men? <lol>

 

[LeonhardEuler]

Hey Math, great post by the way. I read this one last night and it's still buggin' me.

That graph you posted is a shocker. I read the description before opening it up. I fully expected to see two conventional bell curves, with the female curve just a bit taller and more narrow than the men's bell curve. But of course, the men's bell curve is upside down! I guess there's a wealth of genius men and a wealth of drooling morons as well, but few in between. I wonder if any more recent studies have confirmed this 'upside down' bell curve. It seems like one of those things you scratch your head over and say, "That just can't be!" I'd love to know if any more recent studies have been found to support the original one.

~

On a side note, have you noticed that almost all the winners of the Darwin Awards are men? <lol>

That graph is hard to read, if you look carefully you'll notice that it's not supposed to be two bell curves. The x-axis is IQ score, but the y-axis is the percent of people who got that score who were male or female. So, for example, of the people who got a 100 IQ score, about 48% were male and 52% female. The values of the two curves always have to add up to 100%, so whenever one curve goes up, the other goes down.

 

[Q_Goest]

Thanks Leon'!

 

[Gokul43201]

Graph:http://www.edge.org/3rd_culture/debate05/images/pinker.slides/pages/pinker_Page_41.htm
So the guys get more geniuses, but also more idiots. Maybe... but to me, studies that old are a bit "suspect". However, the same phenomenon is apparently observed in SAT scores and another study that is mentioned in their debate.

But I don't want to dwell on measures of intelligence. I'm interested in other traits like height, and why males would vary more than females. Why would natural selection favor this?

This is just the thing you wanted to avoid, so I'm sorry for bringing it up again. That graph doesn't make a whole lot of sense to me (and it doesn't help that it's pretty foggy):

1. It looks like they've inverted the histogram for the boys - the numbers near the mean (~5000) look larger than the numbers at the tails (183, 277, etc.). So, I don't see why the graph has the medial points lower than the extremal points, unless they've inverted it for some visual advantage.

2. The y-axis labels make no sense at all. The explanation says the "the y-axis represents the percentage of each sex in each 5-point band of IQ scores." The numbers we see on the y-axis run from 40% to 60%! With numbers that big, how can you possibly have more than 2 data points per sex? How can 52% of the girls score in the 95-100 band and an additional 52% score in the 100-105 band?

3. If you take the boys-curve and reflect it about the 50% line, it looks like it might exactly match the girls-curve...including the strange dip at IQ 75. What is going on? It sure looks like one curve is the exact reflection of the other!

Or am I being intensely dense and/or blind?

 

[LeonhardEuler]

This is just the thing you wanted to avoid, so I'm sorry for bringing it up again. That graph doesn't make a whole lot of sense to me (and it doesn't help that it's pretty foggy):

1. It looks like they've inverted the histogram for the boys - the numbers near the mean (~5000) look larger than the numbers at the tails (183, 277, etc.). So, I don't see why the graph has the medial points lower than the extremal points, unless they've inverted it for some visual advantage.

2. The y-axis labels make no sense at all. The explanation says the "the y-axis represents the percentage of each sex in each 5-point band of IQ scores." The numbers we see on the y-axis run from 40% to 60%! With numbers that big, how can you possibly have more than 2 data points per sex? How can 52% of the girls score in the 95-100 band and an additional 52% score in the 100-105 band?

3. If you take the boys-curve and reflect it about the 50% line, it looks like it might exactly match the girls-curve...including the strange dip at IQ 75. What is going on? It sure looks like one curve is the exact reflection of the other!

Or am I being intensely dense and/or blind?

See my previous post.

 

[Gokul43201]

See my previous post.

Yes, that explains it. Also, it reveals the ambiguity in the explanation provided below the graph, but mostly, it was just me being dense after all!

 

[Moonbear]

That graph is hard to read

No kidding! It took a lot of staring to make sense of it. What an odd way to present the data. I'm not sure that makes any sense at all unless you start out with equal numbers tested, which if you're testing the entire population of a country, isn't very likely. I can't quite read the numbers to add them up, but just testing more boys than girls, or vice versa, will really bias that sort of method of graphing the data.

 

[Math Is Hard]

I've been looking for a clearer version of that graph - sorry, no luck yet. (Maybe I should email Steven Pinker?) I agree, it's an odd way to present the data. Why not just show two distributions so we could easily see more leptokurtic tails for the boys vs. the girls? And, as Moonbear mentioned, there's an awful lot of ways that data could be problematic. Personally, I have flashbacks of "The Mismeasure of Man" when I consider year the study was done. Not that the researchers were unethical, but the results may have been heavily influenced by the education system that the 11 year olds were in at the time.

Still, I have tried to consider the idea of greater variation to see where it led. I was thinking along the same lines as Leonard as there might be some chromosomal explanation for why males would vary more than females in certain traits, but couldn't quite work out the mechanism. What came to mind was my own bias from a sample of California birds. The females were difficult to tell apart, but the males were not only easy to tell from the females - but also from each other! But that was just a sample.

And Q_Goest, you made a really good point that men are more frequent Darwin Award winners. I need to find some more data, but I do know that men are over-represented in known cases of anti-social personality disorder. APD is associated with risk-seeking behavior (which, of course, can lead to Darwin Award winning), so I can't help but wonder if men are also over-represented in disorders involving extreme cautiousness (OCD perhaps?)

Overall, the "evidence" for greater variation I'm finding as I search on the web is troublesome because I am only finding claims about intelligence, and not other traits. There are so many ways that the measures of intelligence could be biased that I don't think we have good measures yet in this regard. But I really don't know where to look until Chris responds with which cognitive scientists made the statements he was referencing.

I didn't come in here to dismiss this idea, I was just hoping that more information was known about it. It wasn't something we ever learned in biology, so I just had to ask. But if there's not much evidence, well, then it looks like it's just rough speculation.

 

[Moonbear]

Still, I have tried to consider the idea of greater variation to see where it led. I was thinking along the same lines as Leonard as there might be some chromosomal explanation for why males would vary more than females in certain traits, but couldn't quite work out the mechanism. What came to mind was my own bias from a sample of California birds. The females were difficult to tell apart, but the males were not only easy to tell from the females - but also from each other! But that was just a sample.

I wouldn't argue at all that it never happens for any trait. I just don't think it's generalizable. Greater variation also wouldn't necessarily mean that the range overlapped with females. Something very important for maternal behavior would probably be pretty tightly clustered among females, but completely different from males, which might exhibit a wide range of variation for that trait (if they possessed it).

And Q_Goest, you made a really good point that men are more frequent Darwin Award winners. I need to find some more data, but I do know that men are over-represented in known cases of anti-social personality disorder. APD is associated with risk-seeking behavior (which, of course, can lead to Darwin Award winning), so I can't help but wonder if men are also over-represented in disorders involving extreme cautiousness (OCD perhaps?)

With regard to OCD, it seems to depend on the type of obsession.

The purpose of the present study was to investigate gender-related sociodemographic and clinical differences among Turkish patients with obsessive-compulsive disorder (OCD). A total of 169 patients diagnosed with OCD by DSM-III-R or DSM-IV criteria were included in this study. Male (n = 73) and female (n = 96) OCD patients were compared with respect to the demographic variables and the scores obtained from the Hamilton Rating Scale for Anxiety (HRSA), the Hamilton Rating Scale for Depression (HRSD) and the Yale-Brown Obsessive Compulsive Scale (Y-BOCS). We found a significantly earlier age at onset in male patients. No significant difference in terms of HARS, HDRS, and Y-BOCS scores was detected between the two groups. We observed a significantly higher frequency of contamination obsessions in females, and that of aggression and sexual obsessions in males. There was no significant difference in terms of the frequency of compulsions between the two groups. We also found that compulsion severity on obsessions/compulsions was higher in females and comorbidity rates of social phobia and schizophrenia were higher in males. Considering our results in combination with those of other studies, similarities rather than differences in gender-related sociodemographic and clinical characteristics of OCD patients across different populations seem to be present.

Tukel R, Polat A, Genc A, Bozkurt O, Atli H.
Gender-related differences among Turkish patients with obsessive-compulsive disorder.Compr Psychiatry. 2004 Sep-Oct;45(5):362-6.

http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6WCV-4D5FNKH-7&_coverDate=10%2F31%2F2004&_alid=527593580&_rdoc=1&_fmt=&_orig=search&_qd=1&_cdi=6748&_sort=d&view=c&_acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=50da0bc284de6e86ee90d1200c7eb848