Friday, November 30, 2012

Personality Testing

Last night, a friend asked me about the Myers Briggs personality test. This morning, I just got an email from my workplace's Staff College inviting me to attend a session on Myers Briggs personality testing.

The Myers-Briggs Type Inventory is obsolete. It was derived from one person's theory, not data. It had some use in the past, and it for a long time it was the best thing we had, but we have better methods of measuring personality now. Using this system today is like using a Model T car or an Apple II computer.

The proper way to measure personality is to use the Big Five personality traits of Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism (OCEAN). These traits were found by analyzing lots of data with advanced statistical techniques. They are reliable, robust, and valid, and have been shown to correspond with a wide variety of important life outcomes. All social scientists who do real work with human personality traits have known this for over a decade.

There is not much else to say about that, but this is a good opportunity to point out a potential problem with how people think about personalities. The human mind instinctively wants to put things in groups or categories. This is a relic of our paleolithic past where it was very important for a brain to instantly and accurately put something in the category of 'snake' or 'vine'.

But the categorization instinct often leads to mistakes in the modern world, and it definitely leads to mistakes when applied to personality traits. All personality traits that can be measured fall on a continuous scale, and most of them have the 'bell curve' pattern where most people are in the middle of the scale and it is rare for people to be extreme.

This means that it is a mistake to try to classify everyone as an introvert or extrovert in the same way that it is a mistake to classify everyone as smart or dumb. Most people have an average level of intelligence and extraversion and other traits. Differences between people are differences of degree, not kind.

If you must impose categories, then you should not use the two categories of intovert and extrovert. You should use the three categories of introvert, normal, and extrovert, where people only get classified as introvert and extrovert if they are far enough from the average.

No comments: