Tech Companies Say They Can’t Find Good Employees. The Companies May Be the Problem.

The trouble is this: Candidates often get so flustered from the pressure and the mechanics of the interview that their performance is significantly hindered.

Companies that rely on these stressful tests are eliminating potentially qualified candidates—to their own detriment, says Christopher Parnin, assistant professor of computer science at North Carolina State University and senior author of the study, which was published in November. This is particularly unfortunate, Dr. Parnin says, given the availability of other, less stressful and perhaps more appropriate testing metrics.

Relying on technical interviews can also have a disproportionate impact on minorities and other underrepresented groups, and discourage job seekers from applying, Dr. Parnin says, noting that even senior-level employees can be afraid to transfer within their own organizations for fear of failing such tests.

Dr. Parnin and his fellow researchers designed an experiment in which 48 computer-science students were given a problem to solve—one used within the past several months in technical interviews at well-known software companies. Nearly half, 22 of the individuals, were tasked with solving a technical problem in private on a whiteboard. The other students went through the typical interview format, in which they were asked to solve the problem out loud with a proctor present. An experimenter was situated nearby but did not interrupt the participant’s thought process. If a participant asked for clarification, the response was brief.

Participants in both settings wore specialized eye-tracking glasses so researchers could take measurements associated with high cognitive load and stress and compare the data.

The disparities in test results achieved by the two groups were telling. About a third of the participants who were watched solved the problem correctly, compared with about two-thirds of those without proctors.

Participants with proctors present reported feeling nervous, rushed, stressed and unable to concentrate. Participants without proctors reported feeling more at ease, and appreciated being given time on their own to understand the problem and reflect on their solution. Analysis of the eye-tracking data supported those feelings: Participants who were asked to problem-solve in a public setting had slower eye movements, a marker for high stress.

Here’s another interesting takeaway for a male-dominated industry that is increasing efforts to attract more women: Researchers observed that among women with a proctor present, no one successfully solved the problem, whereas all of the women working alone solved the problem.

While the study is based on limited data, Dr. Parnin says the results should, at the very least, lead to further research in this area. Additionally, he says, tech companies should consider alternative testing methods for potential new hires and for employees looking to advance to more senior-level jobs. A shift to testing problem-solving skills in a private setting, for example, could relieve some pressure and allow for more accurate assessment, he says. Such a shift could substantially increase the number of qualified candidates, particularly in traditionally underrepresented groups entering the workforce, he says.

Companies could also drop problem-solving tests as currently offered and instead ask candidates to spend five minutes explaining how they would perform a particular job-related task, Dr. Parnin says. Focusing on communication skills in this way, Dr. Parnin says, can reveal how a candidate thinks.

Ms. Munk is a writer in West Orange, N.J. She can be reached at [email protected].