Correlation between LSAT and SAT, for those who asked Forum

[gcu22]

So it appears many have this question, and while I'm not expert, there's some mild correlation(ie if you are smart you'll do well enough on both), but not a specific correlation.

For example
My SAT:1390(high 95%+tile) 730 verbal(98%)
My LSAT 161(84%tile)

My friend::
SAT:1290
LSAT:164

My other friend
SAT:1020
LSAT:145

and finally a high school buddy
SAT:1420
LSAT:167

From message boards around I've seen similar results. My main issue with the test was time,even after a lot of practice. I could never even come close to finishing sections, so although I only missed 4-5 questions out of the ones I attempted, total, I still only did okay. I had such time issues on the SAT, but not as bad, have alwys been a slow test taker, which hurt me a ton on the LSAT.

Bottom line, appears to be a general correlation, but don't assume(like I did before I started practicing) that just because you did well/poor on the SAT that the LSAT will be the same, it's a different test in many ways.

[magnumalv]

I agree. My SAT in no way correlates to my LSAT:

1520 SAT (800 Verbal/ 720 Math)
163 LSAT

It's much harder to score in the 95+% for the LSAT. I believe a large part of that is due to the testing population. I think it's safe to say that law school applicants are, as a whole, of a higher academic caliber than the entire undergraduate hopeful sample. So it will necessarily be harder to score in the top percentiles.

[xikzhao]

I don't think it correlates well either, but it's not necessary true that the LSAT is harder. My SAT score was relatively weak, but I did really well on the LSAT. The SAT Verbal was (at least when I took it) pretty heavily based on vocab. If you didn't know the word, you were dead. Additionally, after 4 years of college, your critical thinking and reading comp improves, all of which makes the LSAT easier. I think it just depends on the type of thinking that you're most comfortable with. The LSAT, more so than the SAT (though it does too), rewards analytical/sequential thinking.

[dtrossen]

[Freedom]

SAT - 1540 single, 1600 combined.
LSAT - 178

I didn't study much for my sat (was aiming to only get 800 on verbal and get the math part later). I think the most important part of my score is how much I read as a kid. I don't speed read, but I read very fast and it left me a lot of time to finish the lsat.

I personally think the analogies section of the SAT (Now defunct) were most similar to the LSAT. They want you to think a certain way about things and once you figure it out, its just a matter of getting your speed up.

[nicksta83]

SAT 1520 (770 V 750 M)
GRE 1370 (680 V 690 M)
LSAT 174

Yea I'm not sure there's much correlation, and i disagree completely that the LSAT is in some way harder. I've also heard a lot of people comparing the GRE and LSAT ( a friend of mine recently told me the LSAT was harder than the GRE without ever having taken the GRE, of course). Of the the tests I've taken I can say from personal experience that i feel the GRE was hardest, then SAT, then LSAT, but that's just me.

[Freedom]

To be sure, I think there is a very rought (.3-.5) correlation between the verbal part of the SAT and LSAT. Reading speed is the thing that is the most limiting factor on the LSAT, followed by logical capability.

[MLBrandow]

Re: dtrossen

I am an example of terrible SAT and great LSAT:

SAT: 1270 (670 math, 600 verbal)
LSAT: 175

I think I was dumb back in high school since my GMAT and GRE were also high.

-------------------------------


In my case (1360SAT, prop-175-178 LSAT), it isn't really a question of corrlation. I didn't study at all for the SAT. I bought one of those SAT prep books and skimmed it on a 4 hour road trip a week before.

It's more of a question of "how much did you try then" versus "how much are you trying now"...

I gave about a quarter-assed effort in H.S. and ended with a 3.6(5.4weighted) and top 2% of 400 classmates. By contrast, I'm actually applying mysef now (call it maturity), and it's noticably paying off.

The question you must ask yourself is, "Am I giving an acceptable amount of effort?" If you are, and your scores are not where you want them to be, I don't think you should be deluding yourself.

By contrast, if you know that you aren't, you merely need to apply yourself and put in the necessary time, and you will see your scores soar.

[dtrossen]

[rtcg26]

College SAT:

1330 (V: 680, M:650) taken June 1998

ACT: 29 taken June 1998

I had a decent SAT/ACT score but still got rejected to a lot of colleges...LOL!

LSAT: 167 taken Dec. 2003

[lcklein]

I am an example of okay SAT, great GRE 1, okay GRE 2, and not great LSAT:

SAT 1400 (710 V, 690 M)
GRE 1 (780 V, 740 M, 760 Logic)
GRE 2 (690 V, 780 M, 6 Essay)
LSAT (163)

I think I was a better test taker right out of undergrad. Now I am old and out of practice, and my LSAT reflects it.

[dtrossen]

I thought the verbal GRE1 = verbal GRE2 and math GRE1 = math GRE2. If so, that is strange the see the numbers change so much. Perhaps there was a big time difference between test takes?

[lcklein]

I was 8 years older for my second GRE and the first test was paper and the second was a computer test. The first GRE I took, I didn't study at all. The second one, I reviewed the math since I had been in architecture for 8 years. Basically, I get worse at these things as I get older.

[bigtard]

Here's a study that focuses on the upper end of the SAT and LSAT scale: http://www.geocities.com/mobiusnu/satlsat.html

[Freedom]

From the information, scoring lowly doesn't mean you you can't do well, though from dtrossen's story and my own experience, I suspect the low SAT, high LSAT scores are merely from people that did not apply themselves, so it would likely be more strongly correlated among people who studied the same number of hours for both tests or studied a certain minimal amount of time for both test.

I'm a little fuzzy on my stats but I believe R is the measure of strength of the linear correlation between the explanatory and descriptive variable. Interestingly math is more strongly correlated than verbal (R=.54 versus R = .5 approximately), but the two factors in combination are even more strongly correlated at .6.

These are just relative measures, looking at the information spread is more interesting, even though the sample size is not large enough to make any real claims on the correlation but it is interesting. It would be more interesting if we could see how peopel did on each section of the LSAT relative to the math section. I suspect that high verbal scores would score better on reading comprehension and high math scores would score better in general in the logic games.

A .6 correlation is very strong, and if we had more data and controlled for more variables this correlation is likely to increase. I also suspect that since the SAT is a little easier and has lower standards than the LSAT, that those that score 1600's are only "guaranteed" around a mean score of 173-175 as indicated by the correlation. That extra 5 point gap requires additional hard work or natural talent. This makes sense since millions take the SAT and a 1600 is 99percentile, wheras 100-130k people take the LSAT and a 172 is around the top 99 percentile.

[Freedom]

Whoops, I meant the correlation to the formula given which gives the expected value for LSAT scores, which give about the same results per section. So the correlation I'm speaking about would say a 1600 is on average a 173-175 and slope in the same way as given in the data.

[dtrossen]

Freedom,
I have always wanted to know what percentile different scores were on the LSAT. Since you got a 1600 you seem to know what you are talking about. Do you know where a table of percentile vs score is for the old SAT. For the SAT I took (1996 or so) what was the percentile for 1400, 1500, and 1600? Do you know what a 1600 is , i.e. 99.2% or 99.6%, etc. I have always wondered what I would have gotten on the SAT if I went back in time. My GRE, GMAT, and LSAT are all 99%, although I don't think I have hit 99.9% on any test except maybe my GRE.

[Freedom]

The student receives a score report approximately 3-4 weeks after administration of the test, with each section graded on a scale of 200 to 800. In addition to their score, students receive their percentile (i.e. the percentage of other candidates with lower scores). The raw score, or the number of points gained from correct answers and lost due to incorrect answers (ranges from just under 50 to just under 60, depending upon the test), is not included; however, one can easily calculate the raw score from the information provided on the score report. Students may also receive, for an additional fee, the Question and Answer Service, which provides the student's answer, the correct answer to each question, and online resources explaining each question.

The corresponding percentile of each scaled score varies from test to test – for example, in 2003, a scaled score of 800 in both sections of the SAT Reasoning Test corresponded to a percentile of 99.9, while a scaled score of 800 in the SAT Physics Test corresponded to the 94th percentile. The difference in corresponding percentiles reflects the number of students who take the test. Generally speaking, the more popular the test, the higher the percentile corresponding to a scaled score of 800.

Some SAT scores and their corresponding percentiles are below.(Before Writing Test Addition)[1]

1600 - 99.9

1500 - 99

1400 - 96

1300 - 89

1200 - 78

1100 - 63

1030 - 50

1000 - 45

900 - 27

800 - 14

700 - 6

600 - 2

500 - 0.5

400 - 0.04

For SAT:

2001-2002 LSAT Percentile Table
(June 1998 - February 2001)
180
179
178
177
176
175
174
173
172
171
170
169
168
167
166
165
164
163
162
161
160
159
158
157
156
155
154
153
152
151
150
149
148
147
146
145
144
143
142
141
140
139
138
137
136
135
134
133
132
131
130
129
128
127
126
125
124
123
122
121
120
99.98 %
99.96 %
99.91 %
99.86 %
99.78 %
99.67 %
99.53 %
99.32 %
99.07 %
98.69 %
98.21 %
97.55 %
97.03 %
95.92 %
94.84 %
93.49 %
91.91 %
90.00 %
88.18 %
85.74 %
83.07 %
80.60 %
77.43 %
74.18 %
70.74 %
67.14 %
63.25 %
59.29 %
55.22 %
51.49 %
47.26 %
43.20 %
39.33 %
35.60 %
32.17 %
28.40 %
25.46 %
22.19 %
19.59 %
16.92 %
14.56 %
12.44 %
10.63 %
9.01 %
7.54 %
6.29 %
5.39 %
4.25 %
3.49 %
2.93 %
2.38 %
1.82 %
1.48 %
1.19 %
0.93 %
0.77 %
0.60 %
0.47 %
0.37 %
0.34 %
0.00 %

From wikipedia - 1500 = 173. 1600=179. percentile wise it seems.

The LSAT has been rescaled very slightly downwards as more people have takent the test.

In addition, the SAT has been scaled downwards a lot as more people hafve taken it. People make fun of Dubya's 1200 but a 1200 back then would be more like a 1400 today.

Also over a million people take the SAT's each year wheras only 100,000 take the LSAT (that number peaked at about 137k during the bad economy) though it is going down again. So take everything with a grain of salt; the SAT is scaled more favorably for smart, motivated people because so many people that don't want to take it have to take it.

You can miss up to 2 questions on each part of the math or verbal depending on the test and still get a 1600, wheras with the lsat's scoring is much less rigorous.

[AR75]

Freedom, you are right.

[typical1L]

1530 SAT, 173 LSAT.

[typical1L]

Is that snark?

I was only responding to the request for other data.

[Alf14997]

i give a golf clap for Uzu's elaboration of the meaning of the phrase gold clap

now if someone could only tell me what snark is I'd be all set.

[AR75]

i give a golf clap for Uzu's elaboration of the meaning of the phrase gold clap

Au + nongonoccoal urethritis = gold clap

[typical1L]

Very interesting that the online usage has lost its sarcasm... in that case, thanks...

[AR75]

It's not lost. You're cool.