• Andrew Kerr
    Hi, I was interested in your blog post of the 27th September where you discuss the US Northwestern University work where they suggest that personality can be boiled down to four personality types (two primary and two secondary). I've read elsewhere on your site about personality traits.

    Can you explain please the difference between types and traits?


  • Lesley Brindle
    Hi, I'm also interested in this. I thought that there were five traits - CANOE. Does the research now show that there are only two main types of personality?
  • John Berry
    Thanks, both, for posting.

    Unfortunately, psychologists are not definitive when it comes to describing personality. The result is that many commentators get very confused and jump to conclusions. Social media is replete with very errored discussions. So, here goes with a definition.

    Personality is the set of habits, cognitive processes and emotional thought held by a person that predicts how they will behave.

    Personality inventories using the Big-5 theory allow 100,000 discrete personality descriptions. To a reasonable approximation, everyone in Cambridge, for example, has a unique personality. That’s quite nice to know!

    This is an important concept in understanding the difference between traits and types.

    This process of assessing personality using a personality inventory and then getting a detailed prediction is too complicated for many commentators. It’s too complicated for many managers who prefer to simplify personality. The desire to simplify anything complicated is a common today.

    But simple generally means dumbed down.

    If you were to take all the people in Cambridge and give them a personality assessment, you’d get 100,000 personalities. You might get some exact matches, but it would be few. But if you inspected those personalities, you’d find clustering of results. You’d find that people who are towards extroversive are also maybe quite agreeable. You might also find that people who are towards the bottom on the anxious (low neuroticism) scale have something in common with those who are quite conscientious. To get this clustering, you’d have to reduce the fidelity of the data. You’d have to forbid each person to have a unique personality.

    The Northwestern University team used machine learning to do an extreme clustering exercise. They reduced the number of discrete personalities possible by forcing only two possible trait scores (high or low) instead of ten. And they dismissed scores that did not cluster well until they got two clusters with a large number of the remaining modified personalities and two with less personalities. They then declared that personality can be described by four ‘types’.

    The Northwestern University team are quite clear about what they did. It was a machine learning exercise. The results would have low construct validity – it would not mean anything in real life.

    So, a type comprises a large number of personalities that have somewhat similar traits.

    Whilst the Northwestern University work might be dismissed as meaningless, it does however have some academic interest. We may see more on this in the coming years.
  • Jasmine Smith
    Thanks for the explanation. Very comprehensive.
  • John Berry
    Thanks. Happy to help.

Add a Comment

Welcome to our TimelessTime Community - come on in.

You'll benefit from facilitated learning though debate with like-minded peers. You can ask questions. You can share your knowledge by answering questions. You can access our Level 7 Training, and check the book reviews.

And the best part - it's all free.