<p>
</p>
<p>“Easier to predict” ≠ good predictor of.</p>
<p>Wouldn’t you have to define “good” with some sort of regression value? Good is a subjective term and really has no place in a quantitative discussion unless it’s already defined.</p>
<p>
</p>
<p>Good as in… useful?</p>
<p>The fact of the matter is that for most high school seniors and college freshman and juniors, the SAT is probably the most accurate predictor of LSAT scores.</p>
<p>Is it perfectly accurate? Of course not, but it does give a general idea of what someone should aim for on the LSAT before he takes a practice test.</p>
<p>false. he/she should aim for 180 no matter what. i personally scored a 173 when my “predicted score” was 165. i would hate to see someone not live up to their potential on the LSAT because some formula said they should score X</p>
<p>
Yeah, that’s the regression value we’re looking for.</p>
<p>
</p>
<p>According to the equation in post #20?</p>
<p>I’d say the projection is only useful in the following context: after you’ve been studying for a reasonable amount of time, if you are not in the range where you’d like to be, using an SAT prediction helps you decide whether it is reasonable to keep studying and try a retake.</p>
<p>post #20 puts me at 170. i was using the “remove a zero and add a 1 in front to your verbal score” thing</p>
<p>
Sounds like it’s pretty damn accurate to me.</p>
<p>From what I’ve seen the equation in post #20 definitely requires to absolutely max out your potential though. You better be studying, because it’s not going to happen by luck.</p>
<p>All,</p>
<p>Please do not laugh! I think that it is useful for HS students as a guide for career planning. I told my sons that their UG GPA would probably be their HS UW GPA - 0.25. Please comment on this.</p>
<p>I used arklogic’s predictor - "LSAT = (SAT)/20.7 + 100.7 " to list a few numbers:</p>
<p>Student A: SAT (M+R) - 1538 — LSAT ~ 175
Student B: SAT (M+R) - 1435 — LSAT ~ 170
Student C: SAT (M+R) - 1331 — LSAT ~ 165 (I think this may be too high?)</p>
<p>Would you think that these are reasonable correlation? What tier of law school would students A, B, and C have a chance of getting addimission assuming a 3.75 GPA? </p>
<p>Also, would anyone know if there is a predictor for verbal part of the MCAT based upon CR SAT score?</p>
<p>Thanks</p>
<p>
Impossible to say. Depends too heavily on which HS and which UG. For example, kids from my HS universally see their GPAs rise – usually by a LOT – in college unless they go to MIT.</p>
<p>Arklogic’s predictor is very well known and circulates the Internet pretty well. From what I’ve seen it holds up pretty well, anecdotally. With that said, there are enough exceptions to it that I don’t think it should motivate a student to give up law school ideas until they’ve taken several practice LSATs.</p>
<p>Anyway, in your hypothetical, your Student A will be admitted to a top-six program. Student B will be at a top-14 program, and Student C will be at a top-25 or so.</p>
<p>
I’ve never seen one. I’m pretty sure it’s impossible to build because the MCAT’s Verbal section is so much more difficult than the SAT (and, for that matter, the LSAT).</p>
<p>There are, what, 75 questions on the MCAT Verbal? (Somebody double check this.) Missing a single question docks you 1 point. A second question docks you a second point. On most exams, a third question wrong docks you a third point. And yet this percentage scoring is usually within the margin of error on the SATs.</p>
<p>So an 800 on the SAT’s RC could easily translate anywhere from a 12-15 on the MCAT V, which is already a pretty large range. You can imagine that the scaling causes similar problems even when you’re down in the 9-12 range on the MCAT. And anything below a 9 means that you’re not getting into medical school.</p>
<p>BDM,</p>
<p>Thanks for the information. This is a useful 1st order approximation which could shed some light on the possible career paths of a student. However, I am surprised to learn that MCAT verbal is so much more difficult than the LSAT. If it is indeed the case, I fail to understand why a medical doctor needs to have better verbal skill than a lawyer.</p>
<p>It isn’t a matter of need – it’s basic laws of supply. There’s so many fewer medical school spots than law school spots that would-be-medical students get filtered out much more aggressively than would-be law students. The result is that eventual medical students have much higher skills at everything than most eventual law students, because so many more premeds get filtered out during the process.</p>
<p>Because of that, the “spread” on exams can be geared toward the upper ranges.</p>
<p>In other words: medical school admissions screens out eighty percent of the students who try to take the MCAT. That means that the test can use its range to spread out the remaining 20%. Meanwhile, law school admissions screens less aggressively than that. (I don’t know what the number is, but I’m assuming it’s less than 80%.) So the LSAT has to compress them, relatively speaking.</p>
<p>For that reason, the MCAT can be a much harder test – even on skills which might matter more for lawyers.</p>
<p>(For example, there’s essentially no difference between a 175 and a 180 on the LSAT. There aren’t many situations where a kid will get rejected with a 175 but would have done great with just 5 more points. Meanwhile, the difference between a 37 and a 42 on the MCAT is definitely significant.)</p>
<p>I just want to point out I’m not sure I’ve known anyone that these predictors have accurately worked for. I scored a 1280 (M+V) combined on my SAT with no studying (and I only took it once). And just took my first practice LSAT and scored a 167 with no studying. That definitely does not correlate. And should correlate even less after I study and take a class this summer. A friend of mine scored a 1560 (M+V) on his SAT with barely any studying and just scored a 163 on the LSAT this past fall with taking a class and doing a good amount of studying outside of it.</p>
<p><a href=“For%20example,%20there’s%20essentially%20no%20difference%20between%20a%20175%20and%20a%20180%20on%20the%20LSAT.%20There%20aren’t%20many%20situations%20where%20a%20kid%20will%20get%20rejected%20with%20a%20175%20but%20would%20have%20done%20great%20with%20just%205%20more%20points.%20Meanwhile,%20the%20difference%20between%20a%2037%20and%20a%2042%20on%20the%20MCAT%20is%20definitely%20significant.”>quote</a>
[/quote]
</p>
<p>Kind of being a stickler here, but you’re comparing a 5-point difference in a 180-point test vs a 45 point test, so of course 5 points is going to be more significant on the 45-point test. If you were to talk about the same situation with the MCAT instead of the LSAT, where a kid gets rejected with a 37 MCAT, the case probably isn’t that he would have done great with 5 more points, it would likely be due to another factor other than the MCAT since anything 37+ is stellar, and because medical schools weigh test scores less and things like ECs more than law schools do.</p>
<p>Turning in an LSAT with every question wrong still gets you a 120, so it’s a 42 point test (MCAT) compared to a 60 point test (LSAT). I could have adjusted the numbers on that proportion, I suppose.</p>
<p>And I do think that the difference between (say) a 41 and a 37 is significant. I think there are some programs where an identical candidate with a 41 will get in while a 37 won’t, particularly at the upper end.</p>
<p>bdm:</p>
<p>any thoughts on the relative “rigor” of MCAT-V vs. the Lit Subject Test? </p>
<p>(Every year on cc there are kids who post a strong SAT-CR score (680+), a 5 on the AP Lit exam, but a sub-600 Lit score.)</p>
<p>I never took the SAT II Lit – but if I had to speculate, I’d think that you’d want to stick with more g-loaded exams like the SAT for comparisons. I take it the Lit test has a lot of specific knowledge needed?</p>
<p>Re #35: Obviously the formula’s not perfect. Your predicted score comes in at about a 162, so it’s off but not by an astronomical proportion. Obviously your friend’s situation is more of an outlier. But anecdotally it seems to work pretty well for most folks.</p>