Razib Khan One-stop-shopping for all of my content

May 8, 2019

Going beyond the sex “grip strength” binary

Filed under: Sex Difference,Sex Differences — Razib Khan @ 11:20 pm

Jesse Singal brought this piece in Deadspin, She’s Got The Strength, But Who Has The Power?, to my attention.

Some very interesting sections:

When we shove the concept of athletic ability—strength, for instance—into the same black-and-white binary that we try to put gender into, we’re wrong. There is no stark line separating what men can do athletically and what women can. Some women, in fact, are bigger, faster, and stronger than some men. A large data set analyzed for a 2018 study looked at the body composition and endocrine profiles of 689 elite cisgender athletes in various sports. When it came to physical attributes there was complete overlap between the men and women analyzed, McKinnon pointed out. For instance, the shortest person in the data set was male, not female. The lightest male weighed the same as the lightest female. There were men athletes and women athletes who had testosterone levels that hit the top of the chart and the bottom. Simply put, the range of any physical characteristic within a sex, (like, for instance, the six feet of difference between the shortest man in the world and the tallest man) is far greater than the average difference in height between the average man and the average woman (five inches). And elite athletes tend to live at the far ends of these spectra anyway.

USA Powerlifting’s response to transgender athletes is head-spinning. The thing about all this talk equating hormone replacement therapy to doping, and the threat to “biological females,” and the “unfair advantages” of “male puberty,” is that it’s based entirely on social perceptions of gender.

“There’s absolutely no scientific evidence at all that supports their position,” said Rachel McKinnon, an expert on athletes’ rights and a professor of philosophy at the College of Charleston, and a world champion track cyclist to boot.

Recently a very successful person told me that mathematical intelligence is probably overrated in comparison to verbal intelligence. It is true that some women are bigger, faster, and stronger than some men, and therefore a lot of social policy follows from this truth? Well, empirically that seems to be the case today.

Despite the irrefutable sophistication of words, I decided to pull some data from the National Center for Health Statistics. These data are useful because they separate by sex and age (and in some cases race/ethnicity). Rather than focusing on ranges, I was curious about the distributions for two characteristics:

  • Height in males and females of a range of ages and between the sexes
  • Dominant hand grip strength in a range of ages and between the sexes

In some cases, there were age intervals, so I simply took the midpoint (e.g., 25-29 becomes 27). Also, they had an 80 and over category. I just left that as 80.

First, let’s look at the age. The figure below shows the distribution of height for males and females over the years, with intervals along with two standard deviations for each age.

Let’s zoom in on puberty.

None of the results should be surprising. I assume many of you remember age 12 when the girls were taller than the boys?

Now let’s look at the distribution at age 25.

You probably want to know the overlap. The shaded area is 51% of the total area under the union of the two curves.

Not a great surprise. Men are taller than women, on average. But there are many women taller than many men.

But what about grip strength? This is one of those standard metrics that’s used to measure health. People with illnesses tend to weak. So they measure this in many people to get epidemiological data.

Here is the plot by age and sex. Again, two standard deviations. Notice that the two curves are more distinct during adulthood than height. Men have stronger grips than women to a greater extent than their simple size difference.

Let’s zoom in on puberty. Here we can see a very large difference between males and females. The difference is more striking than with height. Between the ages of 12 and 13, males start to zoom away from females in grip strength. There’s basically no difference in grip strength for children of elementary school age, on average.

Now let’s look at the distribution at age 25. Again, it is more striking than that for height.

So finally, what’s the overlap? ~21% of the area under the curves overlaps. This means that are a substantial number of women who could pummel a substantial number of men. But, in these cases, it is the strongest of women and the weakest of men. I would say that that’s not sporting, to be frank.

In media, there is often a depiction of rather petite women taking on larger men in physical fights. Because film and television is fantasy, of course, the women, if they are on the “good” side, will come out victorious (just like the hobbits in Peter Jackson’s films). But reality would not be as pretty.

On the other hand, a large woman and a very small man seem like it would be a reasonably “fair” fight from these data. But I’m certainly glad that Red Sonja was not written and filmed in a more realistic manner, with the towering Brigitte Neilsen avoiding Arnold Schwarzenegger’s Lord Kalidor to only slay smaller fry.

You may have noticed that males are stronger than females when you correct for body size. Why? The average body fat percentage of a male (not overweight or obese, though not fit) is in the low 20s. The average body fat percentage of a similar woman is in the high 20s. In contrast, men have 40% more muscle mass in the upper body, and 33% in the lower body.

I’m just a simple geneticist. I don’t really know much biology. I don’t know what goes on physiologically and structurally beyond what I learned in high school and what I know from being a human being who went through puberty. But something happens that makes males and females quite distinct in their athletic abilities from what I can see in the data.

May 4, 2017

In grip strength a woman in the 90th percentile would be at the 10th percentile for men

Filed under: Grip Strength,Sex Differences,Sports — Razib Khan @ 11:44 pm


In relation to the title of this post, it is more accurate to say that a 35 year old British woman who is in the 90th percentile, at 39 kg, would be at the 10th percentile in 35 year old British men. But she would be at the 50th percentile for a 70 year old man. Another way to look at it is that an average 35 year old woman is as strong as an average 80 year old man.

Also of interest, 90 year old men have stronger grip strength than 10 year old boys (25 kg vs. 17 kgs) but 90 year old women have weaker grip strength than 10 year old girls (14 kg vs. 16 kgs). There is a major difference in life trajectory. Men and women start off with the same upper body strength as boys and girls in elementary school. But between ages 10 and 30 years men really outpace women.

Here’s a chart I constructed from the data with male an female at 10th, 50th, and 90th, percentiles:

Look at what happens to girls/women in the 90th percentile between 10 and 30. Because girls develop faster they are highly competitive with boys up until around 15, and then the “great grip divergence” kicks in. Both men and women get stronger between 20 and 30 (to my surprise), but men gain a lot more. At age 30 in standard deviation units the average women is about two standard deviation units below the average man. This would put the average women at the 2.5th percentile of men assuming a normal distribution. The raw table is below the fold at the bottom of this post. The paper is Grip Strength across the Life Course: Normative Data from Twelve British Studies.

The data to the left are from the paper Hand-grip strength of young men, women and highly trained female athletes. It is a German study, and compared three populations: normal men, normal women, and elite national level athletes in Germany in judo and handball. Judo and handball were selected because grip strength are at a premium in these two sports.

To their surprise the average female athlete was at the 25th percentile for males.

Many people will find this post a bit ridiculous. Who doesn’t know that men aren’t stronger than women?

First, there are some academics who believe that increased training will allow women to converge with men in strength, and therefore they propose to end sex segregation in sports. Proponents of this view say thinks like this:

Could that change? Could women start catching up with men again? After all, people used to say women were unable to handle political office. Even the slowest-converging lines eventually do merge; the truth is nobody knows for sure if that’s the case here. The history of women in sports is a history of being gradually allowed access to social privileges which have made them better athletes, and there could yet be undiscovered factors at play that could make the gap smaller.

There’s no comparison between political office and athletic performance. Second, the slowest converging lines do not eventually merge by necessity.

Second sometimes it is good to have numbers. I am not in the habit of getting into fights with women, ten year olds, or senior citizens, so I don’t have a quantitative grasp of how I “stack up.” I don’t beat my children so I don’t have a good sense of how much stronger than them I am, though my parents did beat me on occasion so I have an intuition about how it is on the other end. My grip strength is probably five times greater than my daughter’s. Good to know.

 

January 5, 2012

The “sex difference factor”?

Filed under: Psychology,Sex Differences — Razib Khan @ 3:10 pm

There’s a new paper in PLoS ONE, The Distance Between Mars and Venus: Measuring Global Sex Differences in Personality*, which suggests that by measuring variation of single observed personality traits researchers are missing larger underlying patterns of difference. The Distance Between Mars and Venus: Measuring Global Sex Differences in Personality:

In conclusion, we believe we made it clear that the true extent of sex differences in human personality has been consistently underestimated. While our current estimate represents a substantial improvement on the existing literature, we urge researchers to replicate this type of analysis with other datasets and different personality measures. An especially critical task will be to compare self-reported personality with observer ratings and other, more objective evaluation methods. Of course, the methodological guidelines presented in this paper can and should be applied to domains of individual differences other than personality, including vocational interests, cognitive abilities, creativity, and so forth. Moreover, the pattern of global sex differences in these domains may help elucidate the meaning and generality of the broad dimension of individual differences known as “masculinity-femininity”…In this way, it will be possible to build a solid foundation for the scientific study of psychological sex differences and their biological and cultural origins.


I’m curious about the reaction of people in psychology to this result. The reason is that I am generally confused or skeptical about measurements of personality difference. I’m not confused or skeptical of differences in personality between individuals or groups. I agree that these exist. I just don’t have a good sense of the informativeness of the measures of difference. People may criticize psychometrics intelligence testing all they want, but at least their methods are relatively clear.

From what I can gather the authors discovered that the differences between sexes on personality were much clearer once you looked for the correlation across numerous single measured traits. This strikes me as similar to what you see in population genetics when you move from variation in one gene across populations to many. While a single gene is not very informative in terms of population differences (e.g., the standard assertion that ~15 percent of variation is between races), synthesizing the variation of many genes allows one to easily distinguish populations, because there is such strong discordance in the correlation of differences. An analogy with traits makes understanding this easy. If you were told that population X tended toward black hair, that would not be very informative. Nor if you were told that population X tended toward straight hair. And what if you were told that population X tended toward light skin? All these traits are common across many different populations. But if you told that population X tended toward straight black hair and light skin, the set of populations which intersect at those three traits together in this direction is far smaller than evaluating on a trait-by-trait basis.

But in regards to the evolution of sex differences there is something that I feel that I can say here. Humans seem to lay between other ape lineages in terms of physical dimorphism. For example, in size the difference between males and females is not as extreme as gorillas, but not as equitable as among gibbons. These differences are traditionally correlated with social structure. Groillas are highly polygynous, and there is a great deal of male-male competition, therefore driving sexual selection. In contrast, gibbons tend toward monogamy (at least in the ideal, as with “monogamous birds” the reality seems to differ from the ideal).

But there is also an evolutionary genetic aspect to sexual dimorphism we must consider: in Genetics and Analysis of Quantitative Traits the authors note that evolution of sex specific traits is not going to occur fast. The reason is simple: aside from the peculiarities on the sex chromosomes males and females are genetically the same. This implies that sex differences on the genetic level may emerge via modulation of gene expression across networks of genes tuned by some “master controllers” associated with differential sex development. All of this added complexity takes time to evolve, with the rough result that sexual differences in trait value take about an order of magnitude longer than other traits to come to the fore. The intuition here is simple: if there is selection for large males, there will be selection for large daughters indirectly. Modifiers which dampen this effect need to emerge, so that sex-specific selection doesn’t have the side effect of dragging the other sex along in terms of trait value (this is a concern when you have traits, such as high testosterone, which might increase fitness in males, but reduce it their daughters). Therefore, if there are sex differences in behavioral tendencies which are biologically rooted (I believe there), they will tend to be universal across human societies and have a very deep evolutionary history.

So that would be the strategy to understand differences in personality across the sexes. Go beyond W.E.I.R.D. populations, as they did in this study. And look for traits where males and females seem to exhibit consistent differences across these range of social environments. I suspect environment does effect the magnitude of differences, but I would be willing to bet money that some differences are going to persist (e.g., inter-personal violence is an area where males will differ due to size and personality).

* I’m really sick of the use of the Mars vs. Venus dichotomy in the scholarship.

October 16, 2010

Female race consciousness as prudence

Filed under: dating,Interracial dating,Racism,Sex Differences,Social Science — Razib Khan @ 3:52 pm

Big Think has a post, Do Women Value Ethnicity Over Income in a Mate?:

The results are striking. An African-American man would have to earn $154,000 more than a white man in order for a white woman to prefer him. A Hispanic man would need to earn $77,000 more than a white man, and Asian man would need, remarkably, an additional $247,000 in additional annual income.

So do women value ethnicity over income in a mate? They certainly seem too. If income was the more important factor in mate choice these numbers would be small; it would take very little additional income to entice a woman to date a man of a different race. The fact that the numbers are so large suggests that a man’s race is significantly more important that his income.

And men? Well the problem is that men don’t seem to care about income at all. So even though their behaviour suggests they care less about their partner’s race than women do, the income needed to encourage them to make the trade-off between races is incalculably large. To really estimate how much men care about race you would have to find a different measure, like perhaps physical beauty.

First, there has been research controlling for physical beauty. So the white male disinclination toward black females can be accounted for mostly by the fact that they aren’t as physically attracted to them. When you limit the sample of black women to those which they are physically attracted to the discrepancy mostly disappears. In contrast, when you similarly constrain the samples of black men which white women judge as attractive the discrepancy in dating preference remains (the same when you do so for Asian men).

All this is not new. I blogged this two years ago, and have gotten bored with the topic (there a regular series of papers which confirm the finding in different circumstances). The sex difference in race preference in the dating literature seems relatively robust. Women care about the race of their partners far more than men, all things equal (in fact, much of the literature suggests men are not concerned about race very much when you control for other background variables). If a site brands itself as “Big Think”, it would be nice to add some value.

I’ll offer a hypothesis in keeping with Ann Althouse’s rule-of-thumb in regards to discussing sex differences in polite company: make sure to make it seem as if women are superior in some fashion. Perhaps women simply have a lower time preference? That is, they’re thinking of long-term consequences. Interracial divorce rates are higher, so women may be making implicit calculations as to the probable success of a relationship as opposed to the short-term benefits of a pairing which men fixate upon. Additionally they may be more liable to “think of the children.” Though I’m generally skeptical of the social science research in this area which indicate that mixed-race children experience stress because of their background, there are plenty of high profile media accounts of people of mixed-race and their “struggles” with their identity. This may shape perceptions of the quality of life of the children. In other words, women aren’t being shallow at all, race is an excellent proxy for all sorts of social-cultural variates which might effect the outcomes of a relationship success, and also the fullness of life which their offspring may experience. Women are then in this model being prudent by using a coarse variate, race, as a proxy for the multi-textured reality of how race is lived in America, and how it matters deeply in the lives of human beings.

To test this sort of model we need data from other societies. There are confounds in this analysis in the USA because Asians, for example, are a small minority who as a matter of necessity can’t really limit their dating pool as much as whites. Additionally, it would be useful to take a fine-grained look at Hispanic dating patterns. About ~50% of Hispanic/Latino Americans identify as white, ~40% as “other”, while ~10% a mix with a substantial number of blacks. The race preference may be mostly a function of perception of cultural values, in which case you’d see that Hispanics don’t exhibit any sex bias in race at all. Then it would not be a matter of women being more racist, but being far less cosmopolitan! Oops, I mean that the low time preference is not operating through a racial proxy but a cultural proxy which is correlated with race. In other words, women are culturally sensitive, while men are culturally insensitive.

March 30, 2010

The sexual straightjacket

Filed under: Biology,Evolution,Evolutionary Genetics,Genetics,Sex,Sex Differences — Razib Khan @ 5:21 pm

Earlier I pointed to the possibility of biophysical constraints and parameters in terms of inheritance shaping the local trajectory of evolution. Today Olivia Judson has a nice post [link fixed] on how the existence of two sexes in many species results in a strange metastable tug-of-war in terms of phenotypic evolution:

In sum, the traits that make a “good” male are often different from those that make a “good” female. (Note: I’m only talking about “good” in evolutionary terms. That means a trait that improves your chance of having surviving offspring.) Since many of these traits have a genetic underpinning, male and female genes are thus being sculpted by different forces.

But — and this is the source of the tension I mentioned — males and females are formed from the same underlying set of genes. After all, in humans, whether you’re a boy or a girl comes down to whether you have a Y chromosome or not: boys do, girls don’t. The rest of the genes occur in both sexes.


The X choromosome in mammals spends about 2/3 of its time in females and 1/3 in males.* And obviously the Y is found only in males. But the rest of the genome is found in both males and females. Judson notes that traits which may be attractive in males may not in females, and which may be attractive in females may not in males. There’s a fair amount of evolutionary psychological work in humans in this vein in regards to the heritability of testosterone and estrogen levels in females and males and how it effects the same and opposite sex (in short, there is suggestive data that “sexy” individuals of one sex, those who exhibit strong secondary sexual characteristics, may be prone to having less sexy offspring of the opposite sex).

Of course you can overcome the balancing tug of war; that’s why you have sexual dimorphism in things like size or facial proportion. But these sorts of traits emerge very slowly because of the equilibrium described above, modifier genes and sex-specific gene expression have to slowly engineer around the overwhelming problem that males and females are genetically no different on a sequence level aside from the Y chromosome. Some estimates put the rate of evolutionary change of sexual dimorphism, that is, trait differences between sexes, between 1 and 2 orders of magnitude slower than conventional population level evolution. Ergo, one would expect that sexual dimorphism differences varying across populations have great time depth, and are probably more interspecific than intraspecific (for example, gorillas vs. humans).

There is naturally a whole field devoted to the study of the origin of sex. But whatever its ultimate rationale and utility an evolutionary context, its existence as a background condition in many taxa may result in a constraint of the exploration of phenotype space, as species divided into two sexes characterized by strong phenotypic differences dance between two sex-specific phenotypic optima.

* Sex determination varies by taxon.

Powered by WordPress