An online monthly research publication by the Ivey Business School 

Volume 13, Number 7
July 2007

Apples and bananas

Mitch Rothstein warns firms using personality tests to take care that they choose the right one

“The evidence is strong that personality assessment can predict job performance, but – and it’s a big ‘but’ – there are all sorts of minefields and swamps you can step into when doing this.”

Professor Mitch Rothstein is talking about the use of personality measures in personnel selection. The practice is now generally accepted throughout North America and Europe. Approximately 40 percent of Fortune 100 companies use personality questionnaires when choosing employees.

Rothstein, who has researched and written extensively in this area, recently published a paper in Human Resource Management Review that evaluates the main trends contributing to the growing use of this practice. “My main message is yes, personality measures add value in personnel selection, but be very cautious,” he says.

Using personality tests to predict job performance has a controversial history. The idea began to catch on in the ‘90s, when a new research technique called meta-analysis clearly showed that the practice had validity. A number of approaches ensued, such as the popular five factor model, which measures extroversion, agreeableness, emotional stability, conscientiousness, and openness to experience.

The Internet is also proving a boost to further growth. Today an applicant can fill out a questionnaire at a computer terminal, and within seconds the report is in the hands of the firm.

The key issue for employers, says Rothstein, is how to choose the right test. Although it is clear that personality tests add value to the hiring process, there are still questions around how particular personality traits match up with certain jobs. For example, it is inappropriate and poor professional practice to use a personality test simply because it purports to assess a characteristic that appears to be desirable. The choice of personality measure must be based on a demonstrated relationship with performance in a specific job. Extroversion may be considered a good quality for team performance because studies show that extroverts are good contributors to teams. But does an extrovert perform better in a function like accounting, where a firm wouldn’t want an employee to be easily distracted by other people?

Rothstein has been involved in an ongoing debate over the value of broad personality predictors versus narrow ones. For example, conscientiousness is a broad personality measure that forms part of the five factor model. Conscientiousness tests for narrower traits such as reliability, neatness, order, and dedication. But it also includes need for achievement. “When you lump need for achievement with reliability, it doesn’t always fit,” says Rothstein. “For example, in a leadership position you would want a person with a high need for achievement, but not necessarily someone who checks the figures over and over again to make sure they are right.”

Using broad rather than narrow measures can be a little like lumping together apples and bananas, says Rothstein. “You end up trying to predict a strange fruit. In the case of a leadership position, what you really want is a measure of bananas – the need for achievement.”

Rothstein believes that there can be value in both broad and narrow approaches, as long as a firm clearly knows what behaviour it wants and picks a measure that predicts it. One way to achieve this is by measuring the traits of existing employees who perform well, and then testing for those traits. “If you don’t pick the right personality measure, you might find yourself measuring something you think that’s important but isn’t,” he says. “And then you’ll end up with people not meeting their performance potential.”

One of the problems for human resource managers is the poor quality of many of the tests on the market. Firms should make sure that the tests they buy are created by qualified psychologists, and backed up by solid research. A poorly designed test, in addition to leading to invalid results, also leaves a firm vulnerable to law suits and human rights challenges.

Another potential problem is “faking.” When filling out personality questionnaires, people often try to make themselves appear socially desirable. Today there are many high quality questionnaires with built-in measures that help pick out people who are trying to make themselves look good. Another effective technique is a warning to users that faking can be detected, and will disqualify them.

Although the science is continually improving, Rothstein cautions HR managers that they should never choose an employee on the basis of a personality assessment alone. “It’s a technique that can increase the likelihood of hiring a high performer, but only as an add-on value to the current selection process.”

Summer Supplement:
Q & A with Professor Deborah Compeau on the impact of IT skills on employee productivity

Previous Issues of Impact

Register for Impact