Selections Interview with Gayle Garrett, Project Director, U.S. News & World Report’s Best Graduate Schools Guide

Page 1 | 2 | 3 | 4 | 5

<-- Previous page | Next page -->

Selections: But then the media reports on how drastically different the results of the major business school rankings are, without really going into why those differences in results exists. How should people be looking at these surveys in comparison? Should the results be compared?
Garrett: I would say that whenever you look at the results of a survey, try to get some information on how the results were constructed, what the model was. That is not always easy. For anyone who sees any statistics whatsoever. . . I know constantly I’m seeing graphs on the front page of really respected newspapers, and I say, “well that is really interesting, but I wonder how big the sample was and exactly how it was chosen, etc.”
Consumers are only going to get more information like that if they ask for it. Because there is constantly going to be the editor who says, “That information doesn’t belong on the front page.” And sometimes it goes on Page 13, and sometimes it gets knocked off Page 13. That is a dynamic process. The consumers have to keep asking for that information.

Selections: How would you convince someone that your rankings are an accurate reflection of a school’s quality?
Garrett: I would say that, first of all, that every single one of the indicators we use is associated with quality. I think people are saying they want to see information at a more specific level, and “your results don’t give that to me.” They are really saying things like, “I want to know how much group work is required in a program, because I consider that to be really important.” That is really not what we are trying to do. We are not answering questions at that level. So, what maybe . . . again, getting back to the consumer and what kind of questions that person wants to ask. Our Web site, besides giving information on our indicators, it gives overall quality, in terms of what peers think about a school. And what these numbers indicate. Every time you look up a school on our Web site, you will see a link to a school’s own Web site. The link is right there, and all you have to do is click on it. So a person can go to a school’s own Web site when looking for the specific information.

Part of the reason that we don’t go into that level of detail is because it then becomes really, really subjective. People say what we are doing is subjective because we are going to deans and recruiters. And they always raise the argument that “you didn’t tell me about X.” If you deconstruct that question, the questioner is actually saying, “I, the questioner, subjectively think that X is the most important piece of information.” If U.S. News were to say that, we would immediately get the criticism of, “Why do you think that is the most important?” That is where the subjectivity comes in. Everyone who is looking for a program should be very clear about his or her own goals and his or her own values, and that is perfectly fine. They should be satisfied. But that particular component of overall quality is not the specific thing that U.S. News should be targeting.

Selections: How much difference really exists between a school that is ranked number 11 and one that is ranked number 9 or number 13?
Garrett: Not a lot. One thing that we did do last year was that wherever there was a tie, we only listed the tied number once. For example, for MBA, there wasn’t a tie until we got to school number 12. But three schools were ranked number 12: NYU, UCLA, and Yale. If you look at the list, there is a difference between a school’s rank and its place level. By which I mean, the rank of a school tells you how many schools are ahead of it. So, the first ranked school has no schools ahead of it. The 26th ranked school has 25 schools ahead of it, but there is no school that is rank number 25. That is because there were three schools tied for rank number 23. I am looking at our top 50 schools. If I start with rank 26, there are nine groupings. Of the schools that are listed there, the 28 schools listed there—of those 28 schools, there are nine separate ranks. Specifically, at rank number 35, you have six schools. If you look over at the next column, they all have scores of 65. If you look at the next three schools that all have scores of 64. . . If one of those three schools with a score of 64 had scores that were slightly different, they would be catapulted from the 41st rank to the 35th. What that is saying is that when people look at the rank number, they should also look at the score, and they should be looking at the full table. Particularly with business schools, there are not wide jumps if you look at the overall scores. Harvard and Stanford were actually tied last year. And this year, Harvard is only one point below Stanford. It is very different if you look at medical schools, where there is a gap between Harvard and Hopkins, and that is because Harvard has such an enormous research budget. It dwarfs everyone else’s. But you don’t have that with regard to business schools. So if business schools are very close in rank, there is not that much difference. So I would say, if any student made a decision between Ann Arbor [University of Michigan] and Dartmouth solely on the basis of that one point, that should not be the basis for making a decision. If the student had gone through all of the various aspects, really looked into the programs, really thought about his or her own goals, and then it was tied. Maybe in that case, yes. But just to say, “Oh, I would never ever go to Dartmouth because it is ranked 11th and Ann Arbor is ranked 10th.” That is really a silly way to make a decision.

Selections: I’m sure you know that the business schools themselves take these rankings very seriously. There are even stories that people will lose jobs because of a slip in the rankings. How do you respond to that?
Garrett: Is it really as bad as some of the anecdotes would make it appear? I would certainly hope it isn’t. If it is, I really would hope that the people who were forcing decisions on that basis can be educated to look a little more closely and clearly as to what is going on. And perhaps to look at school’s performance this year as compared to last year. Not necessarily in terms of where the school is ranked, but in terms of the particular scores it is getting on its indicators. Just as I would say it is silly for a student to base an entire decision solely on one point in the rankings or one point in the score, I think even more if administrations make major decisions only based on that, there is a lot of information in the tables. I have seen schools that have made statements about . . . they point out that the entrance scores for their students have gone up, for example. They should be looking at that kind of thing. I know it is hard for schools, especially if they are just slightly off the page. But any school that provides enough data so that we are able to calculate a ranking has access to that information. We only publish the top 50, but the school itself will be provided its ranking as long as we got enough information so that we could calculate it. And once the school has it, the school is free to publish it. So if a school is just a little bit below the Top 50, they are certainly entitled to publish that information.

Selections: How many schools do you survey?
Garrett: We survey 341 accredited MBA programs. But a large number of those are basically part-time programs. The model really only evaluates full-time programs. Last year, we surveyed 341 schools, but there were only 105 that were full-time programs that gave us all the data—including all of the salary data—that we needed.

We don’t go through and say, “These schools didn’t evaluate other schools, so we are going to knock them out of the rankings.” We don’t do that. What we do in calculating the score for peer assessments is do a trimmed mean. This means that we eliminate the two highest and the two lowest values. And one of the reasons we do that is to take out the effect of people who vote for themselves as a perfect 5.

<-- Previous page | Next page -->

Page 1 | 2 | 3 | 4 | 5