Wednesday, October 26, 2011

Post October 2011 LSAT: PrepTest 64 Scale Score Graphs

The October 2011 LSAT scores were emailed earlier this week, so we hope that you end up with the score you wanted (what should you do now?)! If you didn't, read up on our advice for retaking the LSAT.

Also, if you've somehow managed to get your hands on the PrepTest, you can already analyze your accuracy and timing performance at our online LSAT tracker for free.

For everyone else, you can still benefit from the graphs we've got showing how the raw score translates into the 120-180 scale score and what people mean when they ask "Was it a hard or easy curve?" At Zen of 180, we advocate calling it a generous or ungenerous curve.

All LSATs are normed against the previous so that, statistically speaking, a person who got a 172 on the December 2007 should score within a specified band, +/- 2 of the reported score. This is to normalize performance across LSATs that obviously have different questions: so, the same person might miss 8 questions on one LSAT and 12 on another, but still earn the same scale score of 172. 

The first PrepTest (-8) would be a hard curve, and the second PrepTest (-12) would be a easy curve. Here's where it gets confusing: the latter test would have, on average, harder questions, making the same person miss more and still be performing at the same (scale score) level.

So, at Zen of 180, we refer to these scales as generous or ungenerous. A generous curve makes up for harder questions, while an ungenerous curve compensates for easier questions. The concepts are the same, but the language makes it more clear what the relationship is between the number of questions you can miss and your eventual scale score.
For you visual learners, here's a graph comparing the average and range fluctuations of PrepTests 53-62 in gray bands with the actual range of PrepTest 63 in red.

PrepTest 64 had an odd curve compared to the previous 10 PrepTests: average in the top 1 percent of scores (from ~172-180), and then generous through the bottom quartile (historically so at 158 and 152).

In the graph, the gray vertical lines for each scale score show the range of raw scores that produced that result in PrepTest 53-62. For example, the lowest raw score for 173 was 7 missed questions, and the highest was 10.

PrepTest 64 fit perfectly within the historical average for scale scores 180 through 171. However, 170-150 was a generous cuve, hitting a new record high at 158 and 152, and always well above the middle quartiles, which are shown by the darker horizontal bands cutting across the range lines.
Hopefully this makes it clear why you can take a PrepTest, miss four more questions than you're used to, and yet still get the same scale score. Worse yet is when you miss four fewer questions, and you still get the same scale score and feel like you were tricked by LSAC! Keep in mind that the scores are normalized across all PrepTests, so be sure to compare your scale scores to each other rather than just your missed questions.

Below is the complete graph for PrepTest 63, from 180-120, as well as a graph linking to an explanation of the scale score's interaction with the bell curve.
PrepTest 64 starts out as expected and ends up "generous," even setting new records. Click for larger view.