Your next interview with a shiny new prospective hire might be perfectly pleasant. You might even give her the job. But, as a recent article in the New York Times suggests, that job offer might also be a mistake.

Research, the Times reports, shows that we typically think interviews are extremely important. The article suggests that we not only think the interview is critical and revealing, the center piece of the job applicant process, but that we're really quite good at it. We think that getting to know a candidate gives us a kind of irreplaceable insight that edges out qualifications or past experience.

But in fact, the opposite turns out to be true. What's even stranger? Research demonstrates that even if we know that interviews aren't helpful, we still want to conduct them.

This love of the interview would be fine — except it has a big downside. By relying so much on the face-to-face, you may end up choosing candidates who fit your unconscious expectations and narratives, not the best people for the job. And that means you may well pass over the qualified, valuable, truly right-fit potential employees your team needs.

Here's why the interview blinds you to the real goods out there, and how you can see a little more clearly.

How researchers know interviews stink

Research going back to 1979 suggests that interviewers see what they want to see. In that year the University of Texas Medical School had to increase its class size quickly. Applicants who had been rejected only after their interviews, but who were otherwise qualified, were admitted. By all measures, these students did as well as those who had successful interviews.

More recently, researchers at the Yale School of Management found that interviewers routinely incorrectly predicted the success of interviewees. Worse, interviewers often had more positive impressions even when the interviewees provided random answers to a series of yes/no questions, using a code that would tell the interviewee to answer yes or no.

Here's the thing though: No interviewer noticed that their interviewee was random. And all the interviewers who did end up conducting a random interview felt like they "got to know" the interviewee better! The study suggests rather strongly that the interview process is flatly unreliable.

Why we want interviews (even though they stink)

Short answer, we as a species really like stories. We like to make them up even when the details don't make sense. We especially like to make up stories out of the information in front of us that reinforce our already-held beliefs and perceptions, and we pride ourselves on the ability to make up stories about other people whom we are tasked with judging.

"The key psychological insight here is that people have no trouble turning any information into a coherent narrative...this is true, as in our experiments, when the information is random. People can't help seeing signals, even in noise," writes Jason Dana, an assistant professor of management and marketing at Yale, in the Times.

Before you conduct your next interview

While Dana believes that we'd do well to lose the interview, it's not likely to go anywhere.

Instead, he recommends "structuring interviews so that all candidates receive the same questions, a procedure that has been shown to make interviews more reliable and modestly more predictive of job success. Alternatively, you can use interviews to test job-related skills, rather than idly chatting or asking personal questions."

As you run the interview, you can also be extremely humble about your ability to foresee an applicant's potential success based on one visit. Balance your intuition and impressions with the other data you gather on your prospective hires — and make the data comprehensive enough to trust it. That may just be how you get the team you really want.