Ever wonder how "Kevin" in accounting got his job? Everyone agrees he's a nice enough guy, just not great at what he does.

Chances are an interviewer who had little or no formal training hired him, and likely lobbed softball questions during the interview because Kevin had made a good first impression.

Instead of asking each job applicant the same structured set of questions, the interviewer didn't use a script and made up questions on the fly to support a gut instinct about who the best candidate would be.

"Interviewers who don't have a script might spend half of the interview talking about a sports team or the applicant's hometown, as opposed to discussing how the candidate would fit into this job," says Derek Chapman, an associate professor in industrial and organizational psychology at the University of Calgary.

Derek Chapman

"Research shows that structured interviews will help an employer select the proper applicant eight times more often than an unstructured interview."

Chapman was the lead author in a fascinating study published this fall in the human resources journal Personnel Psychology.

Among its key findings, the study showed that few interviewers actually worked from a script, yet Chapman says 30 years of scientific data show this is a better method to choose the right candidate. And less than three in 10 interviewers surveyed had any formal interview training.

"It's ironic that an interviewer representing your organization, selecting the people who will be part of your organization's future, has little or no training at all," he says. "That astounds me."

David Zweig, the study's co-author, calls the findings disheartening. Not only do companies create bad job fits, he says, but they lose productivity and incur unnecessary costs in time and money dealing with the bad hire and/or in finding a replacement.

Why does it happen? It could be inertia on an organization's part, or company leaders just don't have the time to read the scientific evidence, says Zweig, a professor specializing in organizational behaviour at the University of Toronto. "It's our job to get the message out," he says.

The study data were collected in surveys over two years from 592 interviewers in 500 Canadian and international organizations. Researchers also surveyed 518 applicants before and after their interviews.

Results showed that 28 per cent of interviewers received just over four hours of formal training on average. However, much of that time was spent training people not to ask questions banned by legislation.

Of the interviewers surveyed, those with some formal training were more likely to favour a more structured interview - but not nearly enough do, say Chapman and Zweig.

They define a structured interview as a process where the same questions should be given in the same order by the same interviewer(s) to each applicant. Interviewers need to score each response and take thorough notes because we have poor - and "selective" - memories about what we've heard.

Structured interviews are better predictors than the informal process because asking the same, standard questions provides a fair comparison among all candidates. And the fact is that first impressions, often made in a minute or two, can sway an interviewer.

Something as simple as the interviewer and applicant both liking the Canucks, or sharing the same fashion sense, may tilt the playing field. Interviewers in an unstructured interview don't realize that subconsciously they might be asking easy questions or directing the interview in a way that supports their first impression.

"I've done interviewing, I've done executive assessments," Zweig says. "I know there have been situations when I've looked at a resume (beforehand) and said, 'This guy is good.' Then I go through the structured interview and I realized he's not as good as I thought. I've got hard evidence (in the answers) that becomes hard to ignore."

Chapman adds that structured interviews ask questions that fit the job description, not unproductive generic questions such as: "What's your biggest strength?" or "What's your biggest weakness?" Job applicants are well trained to answer those kinds of questions, he says. And they don't predict how a person will perform in a job.

Instead, interviewers should use hypothetical or behavioural questions. Asking, "What would you do if this happened?" or "Give an example of what you did in a team setting that went off the tracks?" allows the interviewer to probe in greater detail and better predict an applicant's future behaviour.

Chapman also cautions interviewers not to cram too much into an interview. Don't ask two questions to determine someone's personality or ask two questions to determine a candidate's knowledge on tax law. There are valid personality tests or skills tests that can be done separately.

Additionally, he says that in unstructured interviews the interviewer often shifts into recruitment mode before the interview ends.

Again, the interviewer has made up his mind that this is the person for the job without gathering enough information.

Interestingly, applicants in the study who did go through structured interviews felt they were being grilled.Their perception of the company was therefore likely to be negative, a potential downside if the company did want to recruit that person.

The solution, Zweig says, is to tell the applicant in advance that the interview may seem more formal, notes will be taken, but it's being done to ensure that all the right questions are asked.

"It helps prepare people for what to expect. They know everyone is getting the same questions in the same order, in order to be fair to everyone.

"This is a signal to the candidate that this organization knows what it's doing," Zweig says.

And it will eliminate questions down the road. Questions such as, how was 'Kevin' ever hired?

(Mike Dempster can be reached at miked@businessedge.ca, and apologizes to all great accounting employees named Kevin.)