Multiple Measures of Personality: Predicting Job Performance

No single trait can provide a clear picture of the complexity of personality.

Over the last 30 years researchers have debated the value of using personality tests for predicting success on the job. Still, personality testing continues to grow as a primary tool for hiring and promotion decisions in large and small companies. The majority of criticism surrounding personality tests centers on how valid the test results are in an employment setting. In the area of selection and hiring, validity typically refers to the degree to which a test can accurately predict job performance. It is generally agreed that tests producing correlation coefficients of .30 to .50 with job performance are useful. Most researchers have concluded that other types of selection tests often have higher validities than even the most predictive personality measures.

Why then, do personality tests continue to grow in popularity? Their growth can be attributed to a number of factors such as low cost and ease of administration. In addition, personality tests typically show lower levels of adverse or disparate impact on protected classes than other selection tests. In other words, they are less likely to screen out a disproportionate number of women and minorities in an applicant pool. Given these considerations, it is not surprising that researchers have devoted major efforts to improving the validity of personality test usage.

Personality researchers have employed various procedures for improving test validity. First, they have worked to identify specific personality traits highly related to job performance. While a number of traits, such as self-efficacy and locus of control, have shown promising results in this area, there continues to be a great deal of debate concerning which types of traits are the "best predictors" of job performance. Second, researchers have focused on developing new types of personality measures, e.g., tests measuring personality specifically in a job setting. Finally, efforts have targeted measuring aspects of job performance more easily predicted by personality measures. Such measures often focus on behaviors that go beyond basic task performance, such as getting along with others or going "above and beyond" the call of duty.

Each of these methods has produced some positive results. Less research, however, has examined the statistical techniques that are used to determine validity. Nearly all validity studies focus on the relationships between individual predictor measures (i.e., individual personality traits such as extroversion and conscientiousness) and job performance. Using these methods, even the best personality measures are typically found to be less predictive than many other commonly used selection tests such as cognitive ability scales and biodata instruments. When examining how personality tests are typically developed and used in other settings, however, it is not surprising that individual personality measures may serve as poor predictors for job performance.

There is a reason why most personality tests measure more than one trait. No single trait can provide a clear picture of the complexity of personality. Instead, most tests produce scores on five to two dozen personality traits. In addition, clinical interpretation of results from personality tests focuses on using all available data to get the "big picture" regarding a person's personality. The bigger picture approach has only recently been considered in the area of selection and hiring.

A growing number of scientists have begun using different statistical techniques for examining the predictability of personality measures. Methods that utilize multiple scores, such as multivariate regression and the examination of interactions, have shown some very promising results in terms of increasing the validity and overall usefulness of personality tests. For example, researchers have recently discovered that a number of variables may interact with conscientiousness. While conscientiousness alone has been found to be one of the most universally predictive personality measures, its benefits may be even greater when coupled with other variables, such as agreeableness and emotional exhaustion.

In short, researchers are beginning to understand what practitioners have known for years. When dealing with personality measures, you have to consider the big picture rather than scores on individual measures. This new line of research could not only result in the increased validity of personality tests, but could move these tests to the forefront of selection and hiring.