Why the LSAT Isn't Scored on a Curve: Myth and Fact

This post is Part 2 of the "The LSAT Curve" series. The series starts with The LSAT Curve | Test-Equating at LSAC.

Myth: The LSAT is curved solely on how everyone does that day.

A lot of test-takers believe that the LSAT is "curved", meaning that you should try to figure out which month's exam will have the greatest percentage of low-scorers and take it with them.

The idea goes:

-If you take the same LSAT as a lot of lower-scorers, you'll look better than you would have otherwise (by comparison) and get a higher score as a result.

-For this reason, you should sabotage your fellow test-takers. Lace their food with laxatives, steal their prep books in the library, etc. Anything to get a leg up on them.

Unfortunately for the dishonest and sneaky among you, LSAC can't just compare all test-takers who took the February 2010 LSAT with each other and have that be it.

Why?

Perhaps February test-takers don't adequately represent LSAT-takers as a whole.


Fact: Different pools of test-takers might perform differently.

Let's assume for a moment that, on average, February test-takers answer fewer questions correctly than the theoretical "average test-taker" would on any other exam. If this were true, the average test-taker would get a higher score than he/she deserves by taking the LSAT in February (all other things being equal).

LSAC can't allow this to happen. If it did, then a 160 on the February 2010 LSAT would be easier for the average test-taker to achieve than a 160 on the June 2010 LSAT, and the 160s would not, therefore, be "equal." Whether the average test-taker intentionally (and foolishly) took the February LSAT with the purpose of being compared to a lower-scoring pool is less important than the results. LSAT scores would not mean as much. They wouldn't be as reliable because one would have to consider the context in which the exam was taken.

For this reason, and due to the fact that there are minor differences in difficulty between exams, scores are not simply curved based on each "test form" in isolation.

The magic solution?

Test-equating.

***

Read on for Part 3: The Experimental Section and Difficulty of LSAT Questions

Photo by petereed / CC BY-NC 2.0