Can Personality Tests Be Fooled?

By Robert Hogan
Hogan Assessment Systems

Robert Hogan, Ph.D., is author of the popular Hogan Personality Inventory. We thank him for contributing this article to our blog.

To discuss the question of faking on personality measures for employee selection, two distinctions need to be observed. The first is the distinction between measures that are scientifically defensible and measures that are not—all personality tests are not created equal. There are 2500 test publishers in the United States, but only four or five can pass muster in terms of scientific defensibility—the watchword truly is caveat emptor. The second distinction concerns the difference between measures of the “bright side” and measures of the “dark side” of personality.

The bright side concerns factors associated with career success. Most commercial test publishers sell bright-side measures. The problem is that really bad guys, the kind who can ruin organizations—the kind who brought down Enron—all have great looking bright sides. These major trouble makers can only be detected using a competent measure of the dark side. The bright side and the dark side are independent. Dark side tendencies coexist with good social skill, and to the degree that an organization relies on interviews, assessment centers, and bright-side measures to hire employees (and especially managers and executives), they are quite likely to hire people who lie, steal, embezzle, file bogus law suits, alienate subordinates, and cause a wide range of other problems.

Thus, we might say that organizations actually fool themselves by using only bright-side measures, especially those that don’t meet the standards of defensibility.

Then there is the question of whether job candidates can fake these measures. The data are quite clear. When they try, some people can enhance their scores on bright side measures. However, when they try, an equal number of people change their scores in the undesirable direction, so in any broad study of this problem, the results are a wash. Effective faking – in which someone is hired based on his or her test -- is quite hard because people are hired based on a profile, not a single score, and virtually no one can fake an entire profile. Also, the chances of someone successfully deceiving two tests are considerably slimmer than one.

These considerations lead to two conclusions. First, it is quite hard to fake a profile on a well-developed measure of normal personality (the bright side), although it is relatively easy to fake on poorly developed measures of personality—i.e., those containing only a couple of dimensions. Second, people may be able to fake a single scale score, such as a single measure of honesty or service orientation, but no one can fake an entire profile. And it is virtually impossible to fake measures of the dark side, especially because “faking profiles” can be triggered by certain patterns of response. It is, after all, the dark side that causes real problems in organizations.