There are a few things that can kill a business. Some bad publicity. A defective part. A lawsuit.
And the unstructured job interview.
There have been several in-depth studies on unstructured job interviews, and most of them say the same thing: they are bad predictors of performance and are subject to bias. That means the companies who use them are at a disadvantage in the war for talent, which can hamper any organization.
“Interviewers probably over-value unstructured interviews,” University of Pennsylvania Professor Jason Dana wrote in his study about the practice. “Our simple recommendation for those who make screening decisions is not to use them.”
There are several studies that have looked at the effectiveness of unstructured interviews and most agree they are poor predictors of performance. Dana’s study, which was published in 2012, actually found that people could more accurately predict performance if they didn’t conduct an interview at all compared to conducting an unstructured interview.
In the study, Dana had 76 participants predict two students’ GPA for the upcoming school year. For the first student, participants just looked at the student’s background information, including their GPA in prior semesters. For the second student, the participants received all of that information, but conducted an unstructured interview with the student as well.
What Dana found was that the people who just looked at the background information made more accurate predictions than people who had the background information and conducted an unstructured interview.
“The assumption is, if I meet them, I’ll know,” Dana told The Boston Globe. “People are wildly overconfident in their ability to do this from a short meeting.”
But there are more consequences to unstructured interviews than just potentially hiring the wrong people. Other studies have found that they are inherently biased in nature, with hiring managers generally choosing people who are like them.
One study conducted by the American Sociological Association (ASA) found that people were likely to hire people who they would be friends with – or even want to date.
“It is important to note that this does not mean employers are hiring unqualified people,” Lauren Rivera, who conducted the ASA study, said in an ASA press release. “But my findings demonstrate that – in many respects – employers hire in a manner more closely resembling the choice of friends or romantic partners than how one might expect employers to select new workers.”
That’s troubling. Not only is that unintentional bias bad for the company because strong candidates are being excluded, it is fundamentally unfair to the applicants.
Studies show that the solution is using a more standardized hiring process to hire candidates, including using structured interviews where all candidates are asked the same questions. ERE.net, one of the most influential recruiting sites on the web, agrees, suggesting more standardized interviews and focusing in on other data to screen applicants.
The best example of this is Google and its hiring process. Google has a data-rich hiring process that combines both structured interviews and background information to determine the best people for the job. And they constantly analyze their process to ensure it remains effective.
Through this method, Google’s founders have put aside their own biases after they were proven irrelevant. Specifically, Google founders Larry Page and Sergey Brin both originally believed a candidate’s college GPA would be a good predictor of their performance, but when the numbers proved otherwise, the company stopped even looking at it.
“One of the applications of Big Data is giving people the facts, and getting them to understand that their own decision-making is not perfect,” Laszlo Bock, the senior vice president of people operations at Google, recently told the New York Times. “And that in itself causes them to change their behavior.”