Beginning in May 2019, I ran the first iteration of a survey of university professors across the United States about their research interests, their methods, and their individual temperaments. My goal is to measure whether certain personality types tend to cluster into some disciplines more often than other fields; for example, do highly extraverted people tend to go into the social sciences at a higher rate than into the natural sciences? This survey was the pilot test of one an updated version I will run again this year.
The second half of the survey asked 59 questions that measured traits of neuroticism, moral intuitions, and systemizing quotients (SQ) and empathizing quotients (EQ). This article compares the SQ and EQ between 17 social scientists and 17 natural scientists in that survey. The SQ and EQ scales were developed by Sacha Baron-Cohen’s brother, Simon. Dr. Baron-Cohen uses these scales as part of his research on the “extreme male brain” theory of autism. The theory comes from the observations of data that male brains are much more “front-to-back” connected and female brains are more bilaterally connected, accounting for a whole myriad of facts including why female children learn language sooner than males and why males more easily grasp mechanical reasoning or “intuitive physics”.
Baron-Cohen’s research discovered that both male and female children with autism have brains with an exaggerated male profile, leading them to score lower in EQ and higher in SQ. I repurposed these scales in my study to find if there were similar discrepancies between professors in the social sciences and natural sciences. My original hunch was that both fields would have relatively equal SQ scores (because scientists in general are more systematic than the average of the population). I figured the key difference between the two would be that social scientists have above average EQs and that would explain why they are more interested in going into fields studying people. It turns out the reverse was true. Both groups had similar EQ scores and social scientists had significantly below average SQ scores. Natural scientists were relatively balanced with their SQ and EQ being roughly equal.
It's not merely that these results were unexpected, but after accounting for sex differences, my data shows that social scientists have a similar discrepancy between their EQ and SQ scores as people diagnosed with autism, but in the opposite direction. Here is a breakdown of the scores.
Social scientists scored 0.88 in EQ (on a scale from -2.0 to 2.0) and natural scientists scored 0.79. With respect to sex differences within and between the fields, there was virtually no difference between male and female natural scientists (females scored 0.7875 and males scored 0.7889). In the social sciences, this discrepancy was 0.98 for females vs 0.83 for males. As you can see in the table of p values below (the likelihood that these differences are due to random chance), there are no significant differences in EQ scores in any direction as far as the data is concerned. In other words, most professors score about the same in EQ, regardless of sex or discipline.
The data gets interesting when we look at SQ, however. Social scientists scored 0.02 in SQ vs 0.56 for natural scientists. This is where sexual differences start to really shape the scores. Females and males in the natural sciences scored 0.29 and 0.81, respectively. This is a similar pattern of SQ scores to what we find between females and males in the general population. And social scientists follow it too, females scoring -0.18 and males scoring 0.38. As you can see in the table of p values below, the differences in SQ between disciplines and between the sexes are significant – except for between females, and we’ll return to this and explain why later.
The most surprising result from all this is that male social scientists scored over twice as high in EQ than SQ, and their SQ scores were below what is typical of males. Males in the general population tend to score a third of a standard deviation above in SQ and score lower in EQ. This sample of males is the only that I can find that inverts this trend. I am going to add more questions to these instruments in the next few iterations of this survey and collect larger samples to find out if this trend continues. If it does, what it shows is that this population has an inverted pattern of the “extreme male brain” archetype from Baron-Cohen’s research. Do male social scientists have “female brains”?
I’m interested in finding out because it would add to our understanding of something called the “male variability hypothesis”. The reason I figured to test for sexual differences in my data because of data like this below, taken from the paper “Gender differences in individual variation in academic grades fail to fit expected patterns for STEM” published in Nature in 2018. This illustrates the male variability phenomenon. If you look at any distribution such as IQ or grades, women will have a slightly higher mean score than men. Men, however, have a wider and flatter distribution, which means you will more likely find men occupying the extremes while the average female outperforms the average male.
We should then, if this model is consistent across many more domains, expect to find a similar sort of male variability in SQ and EQ scores. What I found surprising in my data was that I wouldn’t have originally expected the specific extreme of low SQ and high EQ in men to consistently pool up in the social sciences. As I said above, I predicted the reverse – consistent SQ scores across the board with variability in EQ predicting which fields people would fall into.
The next iteration of my survey will cover 4 traits including SQ and EQ, the other two being extraversion and neuroticism. I also added a questionnaire on three factors of disgust (sexual, hygienic, and moral disgust) to measure whether similar discrepancies show up in these across the disciplines. If you would like to support this research, considering sponsoring my content here on read.cash. My long-term goal is to demonstrate how quality independent science and research can be done without being affiliated with a university or taking on unmanageable amounts of debt thanks to blockchain technology, decentralization, and affordable information technology.
