Topics >  Leadership and Managing >  Human Resources >  How to Interview > 

The New Science of Hiring

.2 Correlation between conventional interviewing and successful hiring

As psychologists have pointed out, traditional interviews produce a subjective, acutely narrow view of a job candidate. That view is likely biased--studies have shown interviewers tend to prefer candidates similar to them, judge candidates on fewer criteria than they think they're judging them on, and tend to let biases about matters like race and gender get in the way. "Everybody thinks they're much better interviewers than they are," says Ben Dattner, a New York City industrial and organizational psychologist.

Still, the interview is a brilliant tool if you make certain changes to it. Behavioral interviews have almost triple the correlation of conventional interviews with job success. (To gauge if a hire is successful, academics use measures like the dollar value of an employee's contribution to the company, his or her relative share in overall output, and later performance reviews, promotions, and raises.) Behavioral interviewing involves, by definition, a group of interviewers defining qualities needed for a job, asking candidates to give past examples of how they've demonstrated those qualities, asking the same questions of each candidate, and taking notes throughout. The premise is that what someone has done in past jobs is a superior indicator of what he or she will do in future jobs. It's the same idea behind checking references.

To see how structured interviews work, take a look at Hope Lumber & Supply, where HR chief Bill Vogt credits much of his company's growth to behavioral interviewing. Hope, which is based in Tulsa, brings in $1.2 billion a year selling building supplies to contractors. Eight years ago, when the company was making a fifth of that, Vogt and the owners predicted, correctly, that the housing market was about to surge. If they hired the right managers, they could ride that wave.

Following behavioral-interviewing maxims, Vogt starts by talking to people intimate with the job and deciding what qualities are necessary for it. He has a standard template for what he wants in managers: leadership, a drive to make money for the company and for themselves, ambition, and past operational responsibility. Depending on the challenges of the specific business unit, he'll alter the template.

Then he comes up with open-ended questions that get at the desired qualities. Behavioral interviews use questions that are rooted in the past--"Tell me about a time when"--rather than hypotheticals--"What would you do if?" Vogt digs deep into his candidates' work experience. "I get into the current operation," he says. "What did you inherit? What were the sales margins, accounts payable, percent current status, inventory like? What did you do with that, what did you achieve? Clearly, we're looking for achievers and winners and people very knowledgeable of their operation." Specific questions like these, in addition to assessing candidates' skills, combat resume fraud--it's pretty difficult to lie about sales margins and inventory turns.

Ideally, a team of people will meet with the candidate. That minimizes the importance of any one person's reaction, good or bad. Vogt arranges a panel interview for general questions, and then sets up one-on-one interviews focused on specific areas. Vogt asks about EEOC compliance and OSHA incidents; the CFO asks about accounting details; the COO asks logistics questions. In any behavioral interview, questions should be job-related, to keep the interview relevant and to avoid discrimination complaints. To the extent possible, every candidate should be asked the same questions. Interviewers should take notes, and should get together to discuss their views just after the candidate leaves.

Step 2

In which the candidate relives college-entrance tests

As helpful as behavioral interviews are, they're even more effective when combined with employment tests, many of which are now administered online. These are given to candidates to assess either cognitive abilities (cognitive tests are filled with SAT-like verbal and math questions) or personality traits (personality tests include preferential questions like "Would you rather spend a night at home alone than go to a crowded party?" or biographical questions like "Were you a class officer in high school?"). While cognitive tests have a slightly closer correlation with job success, personality tests are useful both as a basis for interview questions and for subsequent development. For the best results, companies should use both sorts of tests or a single test that combines the two elements. (For a roster of tests, see "Choose Your Weapon".)