Predicting LSAT from SAT?

<p>Are the scores related? Top 0.5% to top 0.5%?</p>

<p>The usual formula, which is very imprecise, is (Math+Verbal)/101 + 21. It seems to be approximately right in most cases, but obviously there is a large standard deviation.</p>

<p>Thanks Mike. I think you are on the right track but the results with your formula do not seem to work. This formula: (SAT/20)+100 produces a 180 for a 1600 Math plus verbal SAT, so I guess I am answering my own question. </p>

<p>This is purely hypothetical on my part. Does anyone have more information?</p>

<p>I know about half a dozen students who scored 1600s on the SAT and later took the LSAT. Only one got a 180 on the LSAT. (Others all scored from 176-179.) So, I certainly would not assume that getting a 1600 on the SAT is as difficult as getting a 180 on the LSAT.</p>

<p>[url=<a href=“http://lsatblog.blogspot.com/2009/02/sat-lsat-correlation-predict-scores.html]This[/url”>Correlation Between SAT and LSAT Scores?]This[/url</a>] should answer your questions about SAT/LSAT.</p>

<p>That’s erroneous reasoning from you. As jonri points out, there’s some recentering that needs to be done around the top edges. A perfect score on the SAT correctly does not predict a perfect score on the LSAT.</p>

<p>I think it is useful to understand that there is, at least, a mild correlation. Agreed that there are no guarantees and preparation is essential.</p>

<p>Just a slight correction to bdm’s post - His formula should be (M+V)/21 + 101.</p>

<p>What he posted is clearly a typo as, among other things, it maps 1600 to about a 37.</p>

<p>Note that it is no mistake that 1600 still maps to less than 180. The LSAT spreads the top of the distribution more than the SAT, so, even in a broad predictive sense, the group that scores 180 should only map to some reasonably small subset of the group that scores 1600.</p>

<p>My apologies; PABitz is right on this one.</p>

<p>not a subset…</p>

<p>FWIW, I think that the verbal SAT score should be given more weight in the formula than the math score, since there is no math on the LSAT.</p>

<p>Statistical correlations tend to show the exact opposite, actually. Math tends to be MORE correlated than Verbal.</p>

<p>This actually makes sense; if you sit down and look at an LSAT, it’s about 3/4 mathematical and about 1/4 verbal.</p>

<p>“not a subset…”</p>

<p>True, but the point I was trying to make was a limited one about 1600 ~/~ 180 (math symbols don’t work very well on this board). If it isn’t even the case that 1600 ~ 180 if you assume perfect correlation of the underlying scoring tendencies (which would produce a situation in which the 180 group is a subset of the 1600 group), then it certainly won’t map once you relax the clearly fallacious assumption re:perfect correlation.</p>

<p>Results can vary lots depending on how much you study. I’m betting most people who virtually didn’t study for the SATs but then studied significantly for the LSATs would score relatively better on the LSATs than on the SATs.</p>

<p>

</p>

<p>Sure it is. That’s why there’s no math in it.</p>

<p>I can rephrase if that’ll make it easier to understand. About 75% of the LSAT uses skills which are similar to the SAT Math; about 25% uses skills which are similar to the SAT Verbal.</p>

<p>Bluedevil, I’m having a hard time accepting your claims. Given that there shouldn’t be any correlation between math and the RC section, your reasoning suggests that both LG and LR utilize solely mathematical skills. </p>

<p>Any problems I have encountered in LR are more with understanding nuances of stimuluses and more of a comprehension-type conflict as opposed to certain misunderstandings that could arise out of mathematical components (i.e percentage vs. absolute numbers).</p>

<p>It’s not mathematical concepts per se; it’s types of thinking that are much more strongly correlated with mathematical reasoning than with essentially verbal reasoning (contextual reading comprehension, tone-sensing, etc.).</p>

<p>And I think there’s some correlation between math and the RC section; not a lot, admittedly, but enough that I don’t think I’m claiming “solely”.</p>

<p>Can you point me to the data that these claims are based off of? I’d like to see if this can help me rectify some pitfalls I’ve been experiencing with a few LR types.</p>

<p>I can concede that certain things like formal logic or cause and effect relationships can be figured out more easily by people with strong math reasoning skills. But the possible link between mathematical reasoning and RC or a large number of other LR problems seems ambiguous at best.</p>

<p>Well, I’d simply say that there’s similarly a correlation between the SAT’s Math and Verbal sections, too, and similarly there’s probably a correlation between the SAT’s Math and the LSAT’s Reading Comprehension. The data is around somewhere; I’ll see if I can dig it up tonight.</p>

<p>Formal logic and cause and effect relationships basically are the logic games (arrangement of variables according to certain rules) and the arguments section; I forget which one is tagged LR and which one is tagged analysis or whatever the word is.</p>