After this initial committee review, what happens?
For the rest of the pool, we will begin the substantive review. Everyone gets reviewed twice by different members of the admissions committee. In the first half of the round, we have what we essentially call the interview committee. After we’ve reviewed a candidate’s application, we will bring them to committee to discuss who to invite to interview. We try my mid-round to have that process completed so we can have most of those invitations out because we don’t want to jam people or ourselves. But there are times when we go a little bit past that, given volumes and workloads.
Once people are interviewed and the interview reports are in, we reconvene as an admissions committee to make decisions. A lot of times we are seeing people twice in committee: once for an interview decision and then again for the ultimate decision.
And who does your interviews?
They are conducted primarily by trained second-year students, but the admissions committee does some as well. We do probably a quarter or fewer of the interviews.
How many people do you interview in an admissions season?
It’s probably a quarter to a third of the pool. Contrary to other schools, I try to interview fewer rather than more. My personal view is that admissions officers tend to overemphasize the interview and put too much weight on it. It’s not to say that it is unimportant. Otherwise, we wouldn’t do them. But I think it has to be viewed as one element of the application just like the academic record, the test scores, or work experience.
That’s quite different because increasingly admissions officials are putting more weight on the interview because they are concerned that the essays are being polished by consultants and many recommendations may be written by the candidates themselves.
I don’t want to overstate this. The interview is important and it can sway us. We look at it closely. But I think we have a good balanced approach. I don’t think we overweight or underweight the interview. My view is shaped from talking to David Caruso (the special assistant to the dean at Yale College and an expert in emotional intelligence) and (Yale University President and psychologist) Peter Salovey who have done a lot of research in this area. The research is pretty clear that those one-on-one interviews have limited predictive validity, certainly when they are unstructured but even when they are structured. We’ve made ours more structured to heighten the predictive validity but to a large degree the evaluation is as much about the interviewer and where his head is as opposed to the interviewee. So we don’t want to overly weigh that. We don’t think that a 30-minute conversation should overwhelm a number of years of work experience or four years of an undergraduate transcript. It can inform a decision but my view is that it shouldn’t rule the day.
Do your reviewers rate each applicant and recommend an outcome?
We have a rating system that requires an admissions committee member to rate a candidate along certain dimensions. And then there is an ability to identify a suggested outcome: admit, waitlist or deny. I emphasize that that is optional. You don’t have to put down an outcome because what happens at committee is that we have a conversation. My personal view is that it can be helpful to have a signal about where people want to come out but sometimes that can frame the conversation a little too strongly. I want to have the conversation in an unfiltered kind of way. When an applicant comes to committee, the two readers will be the first to talk and present the profile. Then we have a conversation among the entire committee and we use a consensus decision making model. That means that everyone has to agree on the outcome. Typically, the other committee members who didn’t do the review will ask questions of the readers to get at what they didn’t see before making a decision.
How is it decided who reads what?
It is largely random. You can just pull files to read. We sometimes break it up to have one reader look at a specific region, like Europe, China or South America so there is someone with a sense of the overall pool from that part of the world. But then the second reader is always going to be random. And sometimes we break it up by industry.
So is there a vote in committee?
No. There is no horse trading. It’s very much a consensus decision. Everyone has to agree on the outcome.
What if only one person on the committee disagrees?
We just have to talk it out. You have to approach it with the right mindset. If done wrong, it can evolve into you take my person and I’ll take yours. We don’t want that to happen. My team has a good attitude about it and it has worked. It’s definitely more time consuming than just going into a room and making a decision by myself. When I started GMAC talked about the value of a consensus decision-making model. I think it leads to better decisions because it is drawing from a better base of opinions and perspectives. No one view will rule the day.
If you had to guess, what percentage of your admits are easy decisions each year? And what percentage would be composed of tough calls and require meaningful discussion in committee?
It’s a good number of people who require some amount of discussion. There are very few people who we would say are a no-brainer.