"If you made (x) mistakes then you'd likely score (y)"-chart

User avatar
Lyov Myshkin
Posts: 232
Joined: Sat Jun 28, 2008 11:28 pm

"If you made (x) mistakes then you'd likely score (y)"-chart

Postby Lyov Myshkin » Mon May 28, 2012 12:49 am

..couldn't think of a snazzier title, unfortunately.

http://www.lsatqa.com/charts/errors_and_scores

i could probably have chosen a better time to share this than 2 weeks before the day of the test.

User avatar
Lyov Myshkin
Posts: 232
Joined: Sat Jun 28, 2008 11:28 pm

Re: "If you made (x) mistakes then you'd likely score (y)"-chart

Postby Lyov Myshkin » Tue May 29, 2012 4:56 pm

didn't want to make another post, and since this is kinda related:

http://www.lsatqa.com/charts/scaling_and_scoring_trends

you can select any lsat scaled score (presumably your target score) and see what the historical trends have been for the error-cutoff for that score. since the data is a bit jagged, our data scientist used a custom smoothing algorithm so that you could see the movement of the plots a bit more clearly.

User avatar
shifty_eyed
Posts: 1934
Joined: Fri Oct 02, 2009 8:09 pm

Re: "If you made (x) mistakes then you'd likely score (y)"-chart

Postby shifty_eyed » Tue May 29, 2012 5:02 pm

Lyov Myshkin wrote:didn't want to make another post, and since this is kinda related:

http://www.lsatqa.com/charts/scaling_and_scoring_trends

you can select any lsat scaled score (presumably your target score) and see what the historical trends have been for the error-cutoff for that score. since the data is a bit jagged, our data scientist used a custom smoothing algorithm so that you could see the movement of the plots a bit more clearly.


Are the dots the actual data points? And did you average the cut off per year? I'm most interested in variations between June and December curves.

User avatar
Lyov Myshkin
Posts: 232
Joined: Sat Jun 28, 2008 11:28 pm

Re: "If you made (x) mistakes then you'd likely score (y)"-chart

Postby Lyov Myshkin » Tue May 29, 2012 5:08 pm

i could make slight adjustments for a june/december graph though

User avatar
Lyov Myshkin
Posts: 232
Joined: Sat Jun 28, 2008 11:28 pm

Re: "If you made (x) mistakes then you'd likely score (y)"-chart

Postby Lyov Myshkin » Wed May 30, 2012 7:52 pm

shifty_eyed wrote:
Lyov Myshkin wrote:didn't want to make another post, and since this is kinda related:

http://www.lsatqa.com/charts/scaling_and_scoring_trends

you can select any lsat scaled score (presumably your target score) and see what the historical trends have been for the error-cutoff for that score. since the data is a bit jagged, our data scientist used a custom smoothing algorithm so that you could see the movement of the plots a bit more clearly.


Are the dots the actual data points? And did you average the cut off per year? I'm most interested in variations between June and December curves.


I've just implemented a filter based on your suggestion, so you can see the data points for each respective test month. It was a pretty insightful idea, actually, while there are less data points (since we're only looking at one test a year now, sometimes not even for cases like february where they're not all disclosed), you get interestingly different results.

And I just realized I hadn't answered your question (about the dots being actual test data points):

They are, under the 'All' filter, an average cutoff based on the tests for that year. However, now that the filtering mechanism is in place, when you're filtering to see what the cutoff has been like for all February tests, they are indeed all individual tests.




Return to “LSAT Prep and Discussion Forum”

Who is online

Users browsing this forum: cherrygalore, CHyde, Lahtso Nuggin and 13 guests