Gender equality 1/2
An established weariness, or how to overstep the double boundary between the sexes
With the collaboration of Jean-Christophe Aubert
I don’t know if you, like myself, are surprised by how difficult it can be to discuss gender issues in everyday conversation. As soon as things start moving beyond generalizations (“Oh men! Oh women!) and ever so slightly toward the differences between or merits of the sexes, it is not uncommon for a sudden stiffness to appear in your counterpart, whether they believe women to be men’s equals, and humph-we-say-it-over-and-over-and-still-need-to-repeat, which makes it hard for them to so much as bear the mention of gender differences, or rather feel there are essential differences — as opposed to “accidental” as defined by Aristotle — between men and women so deeply embedded in the nature of our sexes and our very “substance” that differential treatment between men and women (who sometimes think this way too) is fully justified.
(It makes for a long sentence, but sometimes letting all the prejudices spill out together can help avoid contamination.)
If you are going to cross that line, so fervently watched by both sides, you better have energy to spare. A touch of bravura and a healthy dose of confidence could prove useful for navigating through this ideologically simplified landscape, where you sometimes feel like an anti-feminist for raising the notion that statistically, for example, women have 10% less muscle mass than men, throw a shotput, bow or hammer 10% less far, have better lexical skills and are less adept in spatial tasks. And other times, in the reigning Manicheism, it feels as though we are denying the statistical characteristics of the sexes when we stand up for equal salaries, or gender equality in company boards and university teaching faculties.
If you ask me, it’s madness and overall exhausting. Exhausting to maintain hope for professional, legal, administrative and social equality that is so self-evident that no one understands why we even need to keep talking about it. And it’s no less exhausting to second guess yourself before discussing findings from research in neuroscience, endocrinology or neuroendocrinology, or new discoveries in their chemical aspects or specificities-statistics between men and women.
Here I will emphasize — heavily — that you can’t be too careful with your “statisticallys,” because I know lots of girls who, without having been exposed to abnormal levels of testosterone, run faster than most men I know, and so many other observable or supposed differences. So statistically, there are differences, and science beckons us to discuss them even if only to conclude that they aren’t of any use to us either politically or socially.
Significance and uselessness of statistics
“Naturally, a difference does not disappear because it is statistical, but it does, for example, make it impossible to base public policies on the fact that, on average, men score higher than women on spatial tasks”
J. Balthazart.
Absolute, average and standard deviation
Saying that women are this or that, that they’re not as good as men in certain cases and better in others, is first and foremost confusing an average with an absolute.
Take a woman who’s 1m.62 and a man who’s 1m.75. I can easily state that the difference, in terms of height, is absolute. It is still absolute if he is 1m.62 and she is 1m.75: there is no other understanding than what is plain for us to see, no other interpretation and no need to otherwise put things in perspective. However, if I undertake to compare measurements between a group of women and a group of men, my previous certainty disappears because I’m suddenly dealing with the averages of all the women’s heights and all the men’s heights. A difference remains, but it is no longer absolute. What we have now is an average difference that says nothing about the concerned individuals specifically.
This displays just why prejudices are so irritating: they project an average piece of information onto individuals that does not indicate any individual reality about them. When a man snickers at me trying to parallel park in a space the size of a handkerchief, he is exhibiting his own confusion as to average and absolute. And my feeling of triumph when the car slips into that space by a hair should be motivated only by having overcome the difficulty of the task at hand, and not that, as a woman, I was able to do what I did, which would be a manifestation of the same prejudice I am here denouncing.
Not only do averages say little about individuals, but they also conceal truths within the groups examined as a whole. For instance, it’s a known fact that 50% of the planet’s wealth is concentrated in the hands of 1% of its owners. Averaging the wealth of the Earth’s inhabitants would not give us the slightest clue of the poverty of the poorest 50%, as their earnings, on average, are inexcusably inflated by the riches of the richest.
The notion of standard deviation allows us to refine the information provided by averages, which we all instinctively know is a means for synthesizing the values of a group that span both sides of the peak. Anyone who has been to school is surely imagining Gaussian distribution: that big heap of average students with the happy few smart kids on one end and a handful of sorry laggards bringing up the rear.
The extensions on either side of the average peak are known as the “standard deviation” and add some nuancing to our sometimes-obscure average: knowing the average temperature in La Brévine doesn’t tell you how many layers to pile on in winter or peel off during the summer months, with a standard deviation between the seasons exceeding 60 degrees. The curve for our example of wealth distribution amongst the Earth’s inhabitants would look something like this, and might prevent us from jumping to conclusions too quickly:
If temperatures in two chosen cities are similar year-round, their curves would overlap because they have the same average and the same standard deviation. We would have a large amount of “overlapping” data, while differences would be non-significant or even negligible. The “effect size” would be very small, meaning the difference (the size) of effects between the two temperatures would be insignificant.
However, if you try to see what happens when comparing this type of curve for an airconditioned home always maintained at 24 degrees and a refrigerator set to 5 degrees, standard deviations would be very slight given that temperatures in both have little variation. The curves would be similar with little overlapping as their averages are far apart. Thus, the effect size would be quite substantial.
What if effect size could strike a blow against prejudices?
All studies conducted on the cognitive differences between men and women give the same results: the averages are close, standard deviations are comparable and effect size is non-significant.
Thereby, if we were to base a public policy on the (proven) observation that 60% of men and 40% of women are above average for spatial recognition, that would mean overlooking the 40% of men less capable than the 40% of women with high-functioning spatial abilities. So, when Mr. X whistles his admiration for Madame who, rather nonchalantly, manages to snatch up some small object in mid-air (other typical gender war situation), it’s because he is only thinking of the portion of women located near the point where the curves overlap, without any consideration for the great many women more skilled in this domain than the average man, and just as good as the best of men.
This too depicts why prejudices are so very irritating: discontented with projecting an average piece of information on individuals that does not indicate any individual reality about them, they also shine the spotlight on congruent portions of specimens and entirely disregard other portions that are in no way less significant.
Saying that “women are not as good of drivers as men” and “men are not as good of multi-taskers as women” are faulty and deleterious statements, insinuating absolute differences when they are in fact the fruit of averages whose curves significantly overlap and whose differences (in effect sizes) are but minor.
This is another way of saying that passing judgement on a gender as a whole or on individual people modulo statistics builds certainties on sand.
Though they can be extremely useful and provide ample information on all sorts of phenomena, statistics are of little use when it comes to making judgements or developing policies relating to gender, no matter their number.
So, there you have it. Our takeaway, once and for all, is that the effect size (gap in overlaps) between the cognitive capabilities of men and women, as proven by the findings of studies tirelessly conducted in all four corners of the globe, is nil. Period.
