The New Grad Rankings Are on the Way

The ability to crunch our own numbers

Mr. Morse,

Has US News considered releasing the raw data used to compile the law school scores?

Perhaps it is there, but I have not been able to find it.

The methodology information breaks down how you weight data.

Some of the data can be found in descriptions of each school (such as library volumes).

Median LSAT and GPA scores can be found from the ABA.

Still, the expenditures by each school on students, and the "other" expenditure category is difficult (impossible for me so far) to find.

In addition, has US News ever considered releasing the formula? I.E. How does the data get reduced to work within the basic methodological formula? Does US News just multiply library volumes by 0.05?

I imagine one reason it may not be released is the likelihood schools would run the numbers themselves and being ranking out the third and fourth tier.

Still, I was just wondering. It would be nice to know how close or how far Tier 3 schools (such as mine) are from the top 100.

Sincerely,

John Nelson

Bob Morse responds: We have published the methodology in detail at law school academic conferences and it's posted on some websites. The data that is used in the rankings "is" available by combining the U.S. News published data and data that law schools can obtain from the ABA Annual Questionnaire. We standardize the data using "z-scores" before we use weights. There are law bloggers (http://agoraphilia.blogspot.com/ ) that have duplicated our rankings, so how to do the rankings is not a secret. Also, any 3rd or 4th tier school can ask and usnews and we will give them their actual rank.

Ed Tech vs. Ed Psych

I'm so glad to see that Educational Technology is finally listed with the graduate programs review. However, I was frustrated to see it lumped in with EdPsych. While the two are sometimes seen together, this skews the rating results for me, personally as I search for a graduate program, but know that two of the best in the nation aren't even listed in the top few given their EdPsych ratings.

Bob Morse responds: It's a typo that Ed Tech is part of EdPsych. IEdPsych should be listed alone. The closest we have is Vocational/Technical

Educational Technology programs

Dear Mr. Morse:

In the 2010 Gradaute School Ranking, please have a separate category for Educational Technology programs. They are actually quite different from Educational Psychology programs and are usually housed in a completely separate department with a College of Education. Thank you.

Robert Reiser

Bob Morse responds: we will take your point under consideration. As I said earlier, it's a typo that it's part of ED Psych.

Engineering ranking methodology

Mr. Morse,

when you consider the Student to Faculty Ratios (full-time doctoral students to full-time tenured or tenure-track faculty (.075) and full-time master's students to full-time tenured or tenure-track faculty (.0375) in the fall of 2007), is a higher student to faculty ratio more favorable or lower ratio more favorble to a school's ranking? That is not clear in your methdology.

Bob Morse responds: In engineering rankings a higher ratio is better. We are using a higher ratio as a measure of fac ulty productivity.

How you input standardized tests....

Is it true that it is not the LSAT scores that are inputted, but , in fact, the percentile ranks for those scores? If so, which data are you using: The percentiles for the previous 4 administrations of the LSAT?

Also, if you are using percentiles for law, is it the same for all of the other rankings that use standardized tests? It seems if this were the case, there is a diminishing return for schools to attempt to raise their medians, especially at the high end, and a dispoportionate effect for schools with medians closer to the national average. Why is this fair?

Bob Morse responds: It's true that we use the percentiles. It's true at the high end there is diminishing returns. However, when you compare a score of 150 to a 170 i.e. one way is 150/170 = 88%. When you use percentile. it's closer to 50 percentile / 98 percentile which measure the real difference in the two scores. In other words, using percentile measure the real distance between the scores for comparison purposes. We use perecentiles in Best Colleges.

Mimicking the Leiter rankings?

Professor Brian Leiter's law school rankings focus on the academic prestige of publishing faculty that is important to fellow academics who focus on publishing. He recently suggested a series of changes to your methodology that would make USN&WR rankings much close to Professor Leiter's own approach. While his approach is fine for those who care mostly about pretigious publishing, and is therefore a useful niche ranking system for that small corner of academia, I urge you not to follow his suggested approach.

Bob Morse responds: We agree that our rankings need to remain unique and we will be careful if we make changes so they don't become like someone elses. We do find it worthwhile to read other views about what we do, but we are able to put them into proper perspective.