Employers often supplement traditional job interviews with selection tools that assess cognitive ability and judgment. And the testing method managers choose-often developed in collaboration with industrial and organizational (I/O) psychologists-can substantially affect an organization's demographic makeup, noted James L. Outtz, PhD.
Outtz, an I/O psychologist and independent consultant in Washington, D.C., demonstrated how different tests can be biased toward people of different ethnicities. For example, in one study, 172 psychology students took a test that evaluated their responses to hypothetical work conflicts. Whites and Asian Americans performed better than blacks and Hispanics when they answered questions in a multiple-choice format, where each question had one correct answer. Yet on a test where students chose a few responses that, in combination, constituted the best solution, the reverse was true-blacks and Hispanics came out on top.
Findings like this suggest that employers should incorporate several testing methods to counteract response bias and adhere to legal obligations and professional standards, Outtz said. Those who design tests for college admissions and professional licensing should also use a variety of answering modes, he added.
"Every test characteristic affects test performance, whether the test designer intended it to or not," Outtz said.
While test designers aim to create a response system that produces the fewest differences across groups, it does no good if the measure doesn't predict job or collegiate success. For this, Outtz recommends studying how each test method relates to the skills needed for the job.
For I/O psychologists, these findings show how important it is to scrutinize testing methods when developing employment selection tools. By doing this, they can ensure that workplaces hire diverse and well-qualified employees.
"Science, practice and social change are not only compatible-they are inextricably bound," Outtz concluded.
Letters to the Editor
- Send us a letter