Many people see traits as something like innate brain circuits or genetic codes that cause our behaviors.

This is why, for example, companies measure personality traits of job applicants. If someone gets a high score on the trait of “cautious,” the thinking goes, then that person will take fewer risks on the job.

It turns out, however, that scores on personality tests cannot predict what specific actions a certain person will do. In fact, research indicates that individuals can routinely act at many different levels of a trait (Fleeson & Gallagher, 2008). So, even those who get high trait scores on conscientiousness could act carelessly in some contexts. The same applies to other personality traits — even someone who is gregarious in most situations can, in certain circumstances, be shy and reserved.

The way we usually think about and measure traits, then, doesn’t do a very good job of predicting or explaining people’s specific behaviors. But that's just what they are often used to do, especially in work contexts.

What are traits?

If traits are not brain circuits that cause a certain type of behavior in all situations, then what are they? Traits could be thought of as summaries of how a person usually acts. The most popular trait framework (the “big five”) was in fact developed by analyzing personality-descriptive words (De Raad & Mlačić, 2015).

A different way to think of personality is as a set of processes — characteristic ways of perceiving, interpreting, and responding to certain features of situations (Mischel & Shoda, 2008).

That means that to predict or explain behavior, you need to know more than one’s trait scores — you also need to know the circumstances, what features of the circumstances one focuses on, and how they interpret and respond to those features.

Consider this hypothetical scenario: Charles is a manager who is up for a promotion at work. His company would like to help him learn new skills, and change some of the ways he acts, so he will be more successful when he leads more people. Charles’ team members have observed that he often gets aggressive in group settings when someone challenges his ideas, and that causes discord in people. His company informs Charles of this reputation, and then helps him learn to see those circumstances differently and practice a new way of responding. Charles is able to change his behavior, and becomes a successful leader.

In this scenario, we didn’t need to know how Charles scored on any personality trait tests. He and his company learned about certain behaviors and the circumstances where they came up. Then, they worked on improving the behaviors. It doesn’t quite matter if Charles has a trait of “aggressiveness” or “thin-skinned,” he was capable of learning to interpret things differently, and then act differently in response. Since (as the research shows) most people are capable of acting many different ways, most people can probably change key behaviors in this way.

Research supports this approach. Studies suggest that in similar circumstances, people often display the same behaviors (Funder & Colvin, 1991; Mischel & Shoda 2005, Tett & Guterman, 2000). In other words, given situations with similar details (such as a group setting, being challenged), certain behaviors are likely (voice raising, verbal attacking).

To predict, understand, and improve behaviors, scores on personality trait tests aren’t very useful. Instead, analyzing how a person responds to certain situations can be much more effective.

How are personality traits useful?

On average, over a population, and over time, traits predict patterns. For example, higher scores on the trait of conscientiousness are correlated with success in many life domains, including work (Shaffer & Postlethwaite, 2013). That means that across a whole company or team, on average, having more people with high conscientiousness scores will probably help achieve some better outcomes.

But for an individual, a trait score probably will not reliably predict or explain specific behaviors. And behaviors, that others can see and that will affect important outcomes, are usually what we want to optimize.

To optimize behaviors, then, we need to know situations and the important aspects of them. Change the situation, or change how one perceives, interprets, or reacts to its details, and it will be easier to change behaviors — no traits needed.