Can someone good with statistics plz interpret this graph? Forum

[Morkiedoodle]

Forgive my stats ignorance, but it looks from this graph like the LSAT score mean is inversely correlated with curve difficulty (December is the easiest curve, with June and October usually being 2-3 points more difficult). Is this right? If so, does this indicate that the February curve tends to be easier than the December curve?

[Source: http://www.lsac.org/lsacresources/resea ... -10-03.pdf ]

[LSATSCORES2012]

It's hard to draw your conclusions from this, particularly because those who take the test in December and February are likely those who are retaking because they scored worse than they would have liked during their original administration.

[Big Dog]

the only thing that you can say for sure is that those who test in Feb score, on average, 4-5 points lower than those who test in October and June. But as the previous poster notes, we don't know anything about the population of testers. Perhaps the Feb admin has a LOT of 3rd timers, for example, whereas Oct & June get more 170+, who consider themselves one and done.

[dingbat]

It's hard to draw your conclusions from this, particularly because those who take the test in December and February are likely those who are retaking because they scored worse than they would have liked during their original administration.

I just figure there are more lazy people who didn't get around to doing the early tests and are doing it last minute

[Morkiedoodle]

It's hard to draw your conclusions from this, particularly because those who take the test in December and February are likely those who are retaking because they scored worse than they would have liked during their original administration.

But aren't LSATs "equated" or whatever to correct for this?

[dingbat]

It's hard to draw your conclusions from this, particularly because those who take the test in December and February are likely those who are retaking because they scored worse than they would have liked during their original administration.

But aren't LSATs "equated" or whatever to correct for this?

What you're referring to is a technique to ensure no one is advantaged by when they take the test / which test they take. If a test is harder, the same raw score gets a better result. It doesn't mean you score better because the rest of the test takers for that particular test are dumber

[SumStalwart]

Sorry, OP. I don't think that you can draw that conclusion. LSAC's process of equating has made it difficult to determine the reason(s) for the score differentials.

Great information, though; I hadn't found this on LSAC before.