Using Tests to Enhance the Interview Process
By John D. Arnold, Ph.D.

For years, organizations have thought of their screening stages as independent sets of activities. You gather application information, test, interview, check references, etc., and each of these activities is a stand-alone process. Two trends, however, have combined to begin to break down this “silo” approach to screening stages. First, the testing stage increasingly includes biographical inventory and personality testing methods that look almost like interview questions. For example:

1. During high school and/or college, I’ve held the following number of officer positions in organizations:
   1. None
   2. One or two
   3. Three or four
   4. Five or more
2. I enjoy it when I get to do new and different things at work. (Response: Strongly Disagree to Strongly Agree)
3. A person can be an outstanding employee even if he or she is late for work more often than his/her supervisor likes. (Response: Strongly Disagree to Strongly Agree)

The reasons for the popularity of these types of items include their strong predictive power and the fact that candidates react more positively to these types of tests than to less “face valid” personality tests.

The second trend involves the almost-universal use of computers to administer and/or score tests. Computers provide administrative efficiency, as well as the ability to use complex scoring formulas. They also provide the ability to score tests and produce reports “on the spot” while candidates are waiting to be interviewed.

This means that information from the testing stage can be passed on to interviewers for follow-up. For example, say a person was neutral on the third question shown above, but passed the test. Knowing their thinking on when it was OK for an employee to be late could be very interesting as you make a decision on whether to hire the person.

Beyond item-level information, you can also supply scale-level information to interviewers. Say, for example, that the person passed the testing stage, but they scored at the “Marginal Pass” level on Customer Service. You might want to probe especially deep in that area to find out whether there are problems, or whether the candidate just responded in a nontraditional manner to some of the test questions.

There are some types of tests where interview follow-up is not appropriate. Skill tests (e.g., math, reasoning, perceptual accuracy) do not provide useful information for interview follow-up. (There’s very little pay-off in asking, for example, why a person missed a math problem or why their reasoning skills are so limited.) But for tests that contain items that resemble interview questions, the follow-up is useful and straightforward. You can simply ask the candidate about their thinking behind how they responded to an item (surrounding your questions about items of interest with questions about other, non-threatening items). Alternatively, you can have pre-developed questions that are less direct that the computer prints out for interviewers to ask when candidates respond in certain ways to the test.

There’s an argument that providing interviewers with information on how a candidate scored on a test may predispose them to judge the candidate in certain ways. There’s some merit in that, and interviewers need to be trained in how to use the information. In general, however, interviewers naturally treat test information with some skepticism. If you tell them that test responses may show areas of weakness, but also might be the product of how candidates approached the test, you usually find that they are perfectly willing to trust their own judgments (in fact, they may be too willing).

On balance, then, test-interview linkages can provide a powerful tool for enhancing the evaluation of candidates. The testing stage can be used, in effect, to ask the candidate dozens and, often, more than a hundred “interview” questions. The interviewer can then follow-up on their responses to those questions instead of having to start cold when he or she sits down with the candidate. The result is an enormous saving in interviewer time in addition to the enhanced effectiveness of the interview.

John Arnold is president of Polaris Assessment Systems, Inc. He can be reached at ph. (313) 822-0646 or email: jarnold@polaristest.com.