Male/Female Differences in Variability Itself
Rethinking the Statistics Underlying Gender Diffferences
Posted Oct 04, 2014
By Blair B. Dawson and Glenn Geher
Comparison of the sexes is as old as contemplative thought. Do men and women differ in aggression? Intellect? Emotional Expression? Empathy? Health? Taste in movies? Statistically speaking, a rich and important discourse exists, complete with ample data, on the mean scores (or averages) between males and females in any number of domains (such as emotional intelligence (females score higher; Joseph & Newman, 2010)).
This said, another important statistical concept, variability (which refers to how much scores in a population vary from one another), consistently receives less attention when it comes to understanding male/female differences. In fact, our research team (the New Paltz Evolutionary Psychology Lab) has recently uncovered that in humans, sex differences in variability across many behavioral and physical domains ends up being the rule. Based on our examination of dozens of past studies that explore differences between males and females, there is evidence to suggest that while males and females may differ in terms of their statistical means in some domains as personality traits, causes of death, emotional intelligence, academic success, and so forth, a story that is just as compelling pertains to the fact that the sexes consistently differ in statistical variability - with, for a large number of variables, males demonstrating greater variability.
To understand the idea of males showing more variability on a dimension than females, consider the following example: If males show a larger range of scores in emotional intelligence than females do, then males demonstrate more variability on this dimension than females do (while females may still score higher on average compared with males). In this example, males vary more from one another in terms of how emotionally intelligent they are compared with females (who, by this example, vary less from one another compared with males on this dimension).
This finding of males showing more variability than females (across varied dimensions) is in line with a highly accepted biological concept, Bateman’s (1948) Principle, which is the idea that males of many species vary more in reproductive success than do females. Based on Bateman’s seminal research on fertility differences across the sexes in fruit flies, it turns out that males vary much more in the number of mates they acquire and offspring they produce compared with females – and this finding is a function of the fact that, as with humans, female fruit flies provide more necessary parental investment than do male fruitflies.
In humans, female morphology also maps onto higher required levels of parental investment – and extensive research on humans (see Brown, Laland, & Mulder, 2009) shows that, as with fruit flies, males (across various cultures) show great variability in reproductive success (with more males either being shut out of reproduction or siring extremely high numbers of offspring) compared with females. Females are more likely to have a number of offspring close to average (i.e., to not show too much in the way of variability).
Why would males vary more from one another compared with females in reproductive output? Males in species such as ours (characterized by relatively low obligatory male parental costs) can sire a greater number of offspring than can females. Females in species with high parental care often have a greater parental costs (Trivers, 1972). If conception occurs in-utero, fetal development naturally gives the female greater responsibility than the male. This fact leads females to be more likely than males to primarily adopt long-term mating strategies and be relatively choosy in selecting mates. Males historically could choose short-term or long-term mating strategies, as their potential investment is not as great (Buss, 2007). This comparatively greater variability in mating-relevant variables in males as opposed to females is theorized by our research team to have extended into other characteristics as previously listed.
Importantly, the idea presented here about males showing more variability across assorted dimensions does not for a second speak to the issue of superiority of one sex over the other. So, for instance, relatively high male variability shows up in the fact that males have a wider range of causes of death compared with females (particularly when men are in their sexual prime (Kruger, 2008)). Having a wider range of causes of death is not exactly superior in any sense .. And, of course, this all is not to say that a female could not die from hang gliding into a bear’s den ...
In our analysis of the published scientific literature, males show greater variability than do females in such diverse domains as:
· Favored Sexual Activities (Peterson, Geher, and Kaufman, 2011)
· General Cognitive Abilities (Machin & Pekkarinen, 2008)
· Verbal Performance (Hedges & Nowell, 1995)
· Math Performance (Hedges & Nowell, 1995)
· Causes of Death (Kruger, 2008)
· Height (Sauro, 2013)
… and in many more domains as well.
This non-comprehensive list is, thus far, in support of Bateman’s Principle at work in humans. And it shows how strongly interconnected variability in reproductive output is with dimensions that cut across the human body and mind. In combination, these findings serve as a strong example of the ripple effect that evolutionary forces have on each of us in ways we might not often consider.
Males and females may differ from one another on average - but, perhaps just as importantly, males and females seem to differ from one another in terms of how much they vary from other members of their own sex. Future research into the nature of male/female differences in any domain would be wise to incorporate an understanding of male/female differences in variability.
Bateman, A.J. (1948). Intra-sexual selection in Drosophila, Heredity, 2, 349–368.
Buss, D.M. (2007). The evolution of human mating. Acta Psychologica Sinica, 39, 502-512.
Hedges, L. V., & Nowell, A. (1995). Sex differences in mental test scores, variability, and numbers of high-scoring individuals. Science, 269, 41-45
Joseph, D. L., & Newman, D. A. (2010). Emotional intelligence: An integrative meta-analysis and cascading model. Journal of Applied Pyschology, 95, 54-78.
Kruger, D. J. (2008). Human life history variation and sex differences in mortality rates. Journal of Social, Evolutionary, and Cultural Psychology, 2(4), 281.
Machin, S., & Pekkarinen, T. (2008). Global sex differences in test score variability. Science, 322(5906), 1331-1332.
Peterson, A., Geher, G., & Kaufman, S. B. (2011). Predicting preferences for sex acts: Which traits matter and why? Evolutionary Psychology, 9, 371-389.
Sauro, J. (2013). The Normal Distribution. In Usable Stats. Retrieved August 5, 2014, from https://www.usablestats.com/lessons/normal
Trivers, R. (1972). Parental investment and sexual selection. In: Sexual selection and the descent of man 1871-1971 (B. Campbell, ed.). Chicago: Aldine Press.
Note that this article is a summary of a manuscript in preparation being co-authored by ourselves (Blair Dawson and Glenn Geher) along with Gokce Sancak Aydin, Morgan E. Gleason, John M. Montgomery, Nathaniel E. Postol, Andrew R. Shimkus, Christopher S. Tripoli, and Nicole A. Wedberg – we call ourselves the New Paltz Evolutionary Psychology Lab! Special thanks to Associate Director of the New Paltz Evolutionary Studies Program, Dr. Spencer Mass for his insight into this work.