Debunking Conventional Personality Tests
Psychologist Adam Meade dislikes the conventional personality tests typically used in the private sector. He believes these tests, often used for employment testing, can be easily faked.
“Reasonably competent job applicants will agree with items asking about meeting deadlines, regardless of whether this truly describes them or not,” he says.
To counter this, Meade, an associate professor of psychology in the College of Humanities and Social Sciences, has developed a software-based personality test based on response time to computer-prompted stimuli. Building in response time can help ensure that test-takers are telling the truth, Meade says.
“It takes time to think ‘OK, what does the employer want to hear?’ When you force people to respond quickly, they have less time to be dishonest.”
How It Works
Meade’s study subjects –students in his introductory psychology courses– take both his test, which is one of seven recipients of this year’s Chancellor’s Innovation Fund (CIF) awards, and a traditional test under two different conditions. In one, test takers are asked to respond honestly, while in the other test takers are asked to make themselves look as good as possible.
The subjects are asked questions gauging the five traits psychologists use to describe personality: openness to change and experience; conscientiousness; extraversion; agreeableness; and neuroticism. The results indicated much bigger differences in scores in honest and faking conditions for the traditional test than there were for the reaction time-based test.
“Test takers could influence how attractive they appear to a potential employer about five times more via the traditional test format versus the reaction time format,” Meade says.
The new employment testing tool could have considerable impact in how organizations choose which job applicants to hire. Studies show that about 25 percent of the variation in job performance can be predicted using the best traditional pre-employment tests. This leaves 75 percent of the variance in job performance that can’t be predicted using current approaches.
Meanwhile corporations waste time and money hiring bad apples. It costs hundreds or thousands of dollars to recruit, select and train new personnel, even for the least complex jobs, Meade says.
And employment testing is a big market. A survey conducted by the American Management Association (AMA) found 41 percent of U.S. employers conduct employment testing of job applicants beyond an interview and resume check. The same survey found 13 percent of all U.S. employers – and virtually all Fortune 500 companies – administer personality tests to job applicants.
Currently, Meade and his research team are busy finalizing a Web-based version of his software. Later this fall, Meade will pilot test the commercial viability of his data with a testing company and confirm that it can predict high-achieving employees. By spring, if all goes well, Meade hopes to have licensing agreements in place.
“This could be a new era in employment testing.”