Men & Women: Personality Differences | Discovering Personality with Dr. Jordan B. Peterson

Well, this is lecture 6, and it's on sex differences in personality. The first thing that you should be aware of is the reason I put it in after neuroticism and agreeableness is that the biggest differences in temperament personality between men and women are in agreeableness and neuroticism. That seems to be true cross-culturally.

One of the things you want to know is, look, what's happened is that psychologists have established a relatively theoretical model of personality. I don't believe that there was any bias built into it, or at least that the bias was minimized by the method. The method was to ask people as many questions as you possibly can and use statistical methods by multiple teams to sort out the relationships between the answers.

I thought that Big Five was a pretty ugly theory when I first came across it. It wasn't one that I warmed up to quickly. I would say it probably took between five and ten years before I was convinced by the brute power of the analysis that it had to be taken seriously. As that happened, because the theory was first laid out only linguistically and statistically, there was no neuropsychology, no neurobiology, and no real theory—there was just lexical description.

So it was kind of like a theory in search of a place to exist. Before it was established, we didn't know what it was also capable of predicting. That all got built up around that, say, five to ten-year period. It became more as we started to understand the relationship between the descriptors and the underlying biology. Then the theory became richer, more useful, and also more credible.

That actually had certain benefits—its lack of initial theoretical attractiveness. It also mitigated against bias; if you find a theory attractive for whatever reason, then you're more likely to maybe be biased in favor of evidence that's supported. I really don't believe that the people who developed the Big Five theory were very much affected by that. There are people who built their reputation on this theory, and of course, they would be biased in favor of pushing their career forward.

But peer review, when it's done properly, keeps that nicely under control. The fact that multiple different methods have converged on the same model and that there's pretty good evidence that it remains reasonably stable cross-culturally, along with its association with the child developmental literature, that's all pretty powerful convergent evidence.

Once you have the theory laid out, you can start to use it to ask and answer interesting questions, like: well, are there differences between men and women? That'd be the first pretty simple question. The next question would be, if there are differences, well then why do those differences exist? But you don't have to get to that to begin with; you can just start with: are there differences?

The answer is yes, there are differences, and I'll tell you what they are, what we know so far. By far, the biggest differences are in extraversion, agreeableness, and neuroticism. The biggest difference is actually in agreeableness. The difference is roughly this: if you picked a random man and a random woman out of the crowd and you said, "Who's more aggressive?" and you bet on the woman, you'd be right 40 percent of the time, which is actually quite a lot of the time.

There's more overlap between men and women, and there are differences even in the dimensions where the differences are the most pronounced. If you took a man and a woman out of a crowd and you asked, "Who feels negative emotion?" and you bet on the woman, you'd be right about 60 percent of the time. That's not exactly right, but it's close enough; you get a flavor for it with that degree of estimation.

So, the first thing you note, and this is the first thing to note from the comparisons of men and women, is that men and women are more alike than they are different, especially at the trait level. We did an analysis with the Big Five; I never published this, but I tried to come up with a set of questions that would maximally differentiate men and women.

So, if you didn't know who had done the answering, could you predict by the pattern of answers? If I took the questions that most differentiated, the best I could ever do was about 75 percent classification accuracy. That's probably a bit of an overestimation because the statistical technique I used was predicated on the fact that we already knew which men and which women had answered these questions and so capitalized to some degree on chance.

So, my guess is that if you used personality testing and it was optimal, you could probably get about 70 percent accuracy in classification, maybe 65 percent, something like that. That's not bad, but it's not enough to think of men and women as fundamentally different.