Page 1 of 1
What's the most the curve shifts from one test to the next?
Posted: Tue Oct 26, 2010 6:36 pm
by longdaysjourney
What's the most the curve shifts from one test to the next? Is there a pattern, or a rough rule? Something like "it's never swayed more than +/- 4 from the previous test?"
This has probably been answered already, but I can't find it.
Re: What's the most the curve shifts from one test to the next?
Posted: Tue Oct 26, 2010 6:38 pm
by bk1
http://lsatblog.blogspot.com/2010/03/ls ... score.html
It seems to shift a lot from test to test, though the shift is getting smaller in recent years.
Re: What's the most the curve shifts from one test to the next?
Posted: Tue Oct 26, 2010 6:51 pm
by incompetentia
The last two tests have reintroduced a large amount of variation from test to test. -11 to -14 to -12...but there is only limited evidence that previous tests influence current ones.
Noticed a fluctuation from the '90s that went from -15 to -9. Which means we can probably eliminate curves harsher than -6 and easier than -18...but we could do that anyway.