Standardized test scores trump transcripts

Don't put down that LSAT prep book just yet.

A study of over 200,000 graduate-school-bound students found that entrance exams like the MCAT and GRE are better predictors of future performance in graduate school than college transcripts.

The study was conducted by University of Minnesota psychology professor Nathan Kuncel and was published last week in Science.

But while the study is the most comprehensive one to date, comparing seven different graduate-school admissions tests, school deans say the study doesn't tell them anything they don't already know.

Penn Law Dean Michael Fitts said the correlation between good LSATs - the admissions test required for law school - and success in law school is why the tests have been so important for so long.

"Performance in standardized tests is something you take very seriously," he said.

Test results may also have more complicated indications.

"There are many people who have done well on the LSAT and have not done well in law school and vice versa," Fitts added.

Robert Schwartz, dean of admissions at the University of California, Los Angeles's Law School, echoed these sentiments.

The LSAT is "a good predictor, [but] some people will come in with low LSAT scores and do very well" in law school.

Still, although they say the study points out the obvious, graduate school officials agree that its findings hold true.

At New York University's School of Medicine, for example, admissions officials use the MCAT to help them deal with a rising number of applications that has increased by 300 over the past five years.

"We just don't have the people or the resources to interview 7,500 people," said Joanne McGrath, assistant dean of admissions at NYU School of Medicine.

Beyond school deans, however, some testing officials still have doubts about the study.

Bob Schaffer, director of public affairs for FairTest - an organization that monitors ethics in standardized testing - accused Kuncel's test of having ties to testing companies that swayed his study.

"There is no new research here, they summarized other studies they have done," Schaffer said.

The study, however, is completely unaffiliated with Educational Testing Services, which sponsors, among others, the LSAT and MCAT, Kuncel said.

He added that his research draws only on work done by university professors unaffiliated with ETS.

Comments (8)

The LSAT is a good indicator of success for computer programmers, but not for lawyers.

You've no evidence to support that statement.

[QUOTE id="aa6e001d-c589-44ec-bb88-d2dee8dcb9d7"]You've no evidence to support that statement.[/QUOTE] Do you? That's what I thought.

wait, why should Rafael need evidence to support YOUR statement? why don't you go review the logical reasoning section of the LSAT

[QUOTE id="de1bc02e-7949-44b6-adf7-271a921f448d"]wait, why should Rafael need evidence to support YOUR statement? why don't you go review the logical reasoning section of the LSAT[/QUOTE] Little Debbie jokes, Ben.

The GRE, LSAT, and SAT are all aptitude tests are meant to predict future college performance but they really don't. A better predictor is college g.p.a., g (general intelligence), and motivation. Look at the manuals and find out if the reliability, validty, and norms of these tests are good enough to generalize to everyone.

Rafael II is wrong. Any law school admissions officer, and this article, will agree that the LSAT is a stronger predictor than GPA. That's because GPAs are almost meaningless when they're compared across schools. Our GPAs aren't the same as the GPA's at Podunk Univ., and they shouldn't be considered that way. There's no way to easily measure "motivation," and you can levy the same complaints about the LSAT against "intelligence tests," if you're that kind of person. People who diss the LSAT generally did badly on it. I sympathize, as I had to take the thing without the thousand-dollar prep courses of my Penn peers, but practice makes perfect.

I think that there is an overemphasis on the "numbers" and very little attention on factors such as the ability to write and to conduct research (after all lawyers primarily research and write). I can name many students with great LSAT scores who cannot write at all, and who are very bad at doing research. But I can also give you examples of students who don't have the stellar numbers but who express themselves very well, who do great research and who are great law school students. The correlation between LSAT and first year law school GPA is generally weak (if you had to model first year LSGPA on the LSAT, the amount of variance explained would be quite low. I have seen some statistics that indicate less than half of the variance in the LSGPA is predicted by the LSAT). And this only considers the first year of law school. All in all, the LSAT is not a great predictor of pretty much anything. It only seems like it is because the LSAC has not developed a better test,and for their part the law schools are too lazy to do the work to select applicants based on more relevant criteria. They just want to filter the thousands of applicants as quickly as possible through their grinder. I have talked to many students, and know many frm an academic standpoint. While there is some correlation between a really good LSAT (over 165) t academic performance, thie tends to apply only at the very top. I can't say why this is the case. Perhaps there is some general intelligence factor, "G", that independently correlates with academic performance. However, students who don't have the stellar LSATs tend to run the gamut, from being mediocre students to being great students. There's just too much variation here and too litte is explained by the LSAT. We eliminate many great students by overemphasis on the LSAT, and also include many bad ones.

Comments are closed for this item.