Deciphering Hidden Biases During Interviews
STEVE INSKEEP, HOST:
Benjamin Franklin said the only certain things are death and taxes. Let's add a third thing: Interviews. At many points, starting with school admissions or a new job, you're going to sit down before someone else and answer their questions.
Which is what NPR's Shankar Vedantam is about to do with us because he's got some new research relating to this topic. Hi, Shankar.
SHANKAR VEDANTAM, BYLINE: Hi, Steve.
INSKEEP: And let's begin this interview. What's the new research about?
VEDANTAM: Well, we often have intuitive beliefs about what's going to help us during the course of an interview. So people say its best not to interview right before lunch, because interviewers are going to be hungry and they'll treat you badly. Or it's best not to interview at the end of the day, because interview panelists are going to be tired. We also believe it matters who comes right before us in the interview schedule.
So if you were applying for a job, Steve, would you prefer to follow a really strong candidate? Or would you prefer to come after a really weak candidate?
INSKEEP: You're asking me.
VEDANTAM: I am asking you.
INSKEEP: You know, I don't think I've ever thought about that in my life. I'm sorry to disappoint you. I just try to go in and make an impression. But in any event, you're laying out kind of folk wisdom here. People pass this on and try to game the system if they can.
VEDANTAM: Yeah. So I mean, I've often thought that it's better to come after a really weak candidate because that can make me look better by contrast. And God knows I need that.
VEDANTAM: Now it turns out that my intuition was wrong. I spoke with Uri Simonsohn. He's a management professor at the Wharton School, and along with Francesca Gino at the Harvard Business School, they've tracked who is successful during MBA interviews - massive studies involving 9,000 conducted over 10 years.
INSKEEP: These are people who were interviewing for college Masters Business degree programs.
VEDANTAM: That's exactly right. And what Simonsohn and Gino find is that going after a really strong candidate doesn't hurt you. What does matter is going last in the day. But there's a catch. It can either really hurt you, or it can really help you. Here's Simonsohn.
URI SIMONSOHN: If it was a good day with many good candidates, it's really a bad idea to the last. But if it was a weak day with many bad candidates, it's a really good idea to go last.
VEDANTAM: So it turns out, what matters is not your contrast with the immediate preceding candidate, but your contrast with the last several preceding candidates. And it stems from a bias that actually has a name. It's called the gambler's fallacy. Let me explain it to you.
VEDANTAM: Let's say I have a coin and I toss it up in the air and it comes on heads.
INSKEEP: All right.
VEDANTAM: I toss it again, it comes on heads again.
VEDANTAM: I toss it third time, it comes on heads again. Now assuming it's a fair coin and I'm not cheating, when I toss it up a fourth time, what do you expect is going to happen?
INSKEEP: You would expect, gosh, it's got to come up tails one of these times.
VEDANTAM: Exactly. That's what I would expect too. And this is called a gambler's fallacy, because the truth is the coin has no memory. So heads is actually just as likely the fourth time it's tails, but we have a very strong intuition that the streak has to be broken.
VEDANTAM: Simonsohn thinks this is exactly what's happens with interview panels, that if you're interviewing candidates and the first candidate is really weak and the second is really weak and the third is really weak, you believe, generally, that there should be an equal number of strong and weak candidates. And what they find is that interviewers add the equivalent of two years of job experience to the last candidate in a row, who is weak, in order to break the streak.
INSKEEP: You'll start grading on the curb in order to find a strong candidate.
VEDANTAM: That's exactly what happens. They think this is happening not just with interviews, but it's probably happening with judges and criminal defendants. It's probably happening with loan officers and applicants. Simonsohn thinks it's even happening with movie critics.
SIMONSOHN: If you're a movie critic, and you watch 10 movies, you will almost never find all of them good, even though sometimes that should happen. Just by chance, you watch 10 very good movies in a week and you should rate all of them to be outstanding, because you don't think all movies are good and then you make the wrong leap of thinking well, any set of 10, it can't be that all 10 of them are good.
INSKEEP: OK. So if I'm an interview candidate for a job or a slot in the school, my goal should be to be surrounded by mediocrity, I get that.
INSKEEP: But, if you're interviewing people...
INSKEEP: ...and you want to avoid this bias, this mistake, what do you do?
VEDANTAM: What Simonsohn thinks interview panels should do is have a spreadsheet where they can see all the candidates they've interviewed, not just on that one day, but over several days. And when you step back and actually say yeah, there are an equal number of strong and weak candidates, even though you may have the streaks of four or five really strong or really weak candidates in a row.
INSKEEP: What you said, basically, has a lesson that's broadly applicable to life, I would think: don't get too focused on the last few things that happened and assume there's a pattern there. Look much more broadly.
VEDANTAM: Exactly. And don't assume that the small pattern necessarily has to reflect the large pattern. So if you toss the coin a million times in a row, and it comes down a million times heads, something is seriously wrong. But if you toss it and it comes down five times in a row heads, it's actually not such a big deal.
INSKEEP: Shankar, thanks.
VEDANTAM: Thanks, Steve.
INSKEEP: You can follow him on Twitter @Hidden Brain. You can follow me @nprinskeep and this program @MORNING EDITION.
(SOUNDBITE OF MUSIC)
INSKEEP: It's MORNING EDITION from NPR News. Transcript provided by NPR, Copyright NPR.