Correlation Between SAT and LSAT Scores?

Yes, there is some correlation between them. However, a low SAT score doesn't condemn you to a low LSAT score, and a high SAT score doesn't guarantee you a high LSAT score.

Some factors that limit the usefulness of any formula relating the two:

1. Brain cells killed while partying in college
2. Differences in your preparation for each exam
3. Differences in the nature of each test

However, I've received several requests recently for information about the relationship between SAT and LSAT scores.

This formula suggests with a moderate degree of certainty (note the qualifying statement here), what your LSAT score might be. I haven't seen any information regarding the origins of this formula, so if you have any, please let me know. This formula is not a prediction or guarantee: simply a suggestion.

LSAT = (SAT Math + SAT Verbal)/20.7 + 100.7

If you studied for 100 hours for the SAT, this formula provides a general guideline for how you might do on the LSAT if you study for 100 hours. Remember to take preparation into account. If you studied for the SAT and got a 1400, but your score prior to studying is only a 150, don't be discouraged. This just means that you have to study.

If you underperformed on the SAT but had a killer GPA in college, this means the LSAT is not likely to accurately predict your performance in law school. If this describes you, I recommend you write an addendum noting this.

For all the visual learners and statistics junkies out there, here's a graph displaying the correlation between SAT and LSAT (using a more complicated equation and a set of only 70 individuals). I found it here. It's far from comprehensive, but I'll share the info and let you decide.

SAT LSAT Correlation Graph

SAT Math LSAT Correlation Graph

SAT Verbal LSAT Correlation Graph





15 comments:

  1. What about ACT & LSAT correlation?

    ReplyDelete
    Replies
    1. looking at the graph and the R2 value, it looks like your little formula is a pretty weak predictor

      Delete
  2. Sorry, couldn't find any info on it. If you do, please post a link in the comments!

    ReplyDelete
  3. Rsquare values are less than 0.4 haha

    ReplyDelete
    Replies
    1. assuming you had full data for all people, you could create a predictor for LSAT with all other variables and you'd probably get good results.

      This is also a truncated sample. There are no low SAT scorers or LSAT scorers. This phenomenon results in lower R^2 values.
      https://en.wikipedia.org/wiki/Correlation_and_dependence#Sensitivity_to_the_data_distribution

      Delete
    2. assuming you had full data for all people, you could create a predictor for LSAT with all other variables and you'd probably get good results.

      This is also a truncated sample. There are no low SAT scorers or LSAT scorers. This phenomenon results in lower R^2 values.
      https://en.wikipedia.org/wiki/Correlation_and_dependence#Sensitivity_to_the_data_distribution

      Delete
  4. An LSAT Prep course told me to take your SAT Verbal score and drop the zero (0) at the end and place a 1 in front of it and then add two (2) which should give you your minimum LSAT capability...

    Example:

    SAT Verbal Score = 560 (drop zero and add 1 in front) = 156 + 2
    LSAT Score = 158

    ReplyDelete
    Replies
    1. actually works pretty well for me, sat verbal 630 converts to 165 LSAT, actual 164

      Delete
    2. Well I got 750 on my sat verbal, and a 177 on the LSAT would be a dream come true, but I don't know, I feel really doubtful about that actually happening in September....and using the above formula in the article, I'm predicted to get a 163. I guess we shall see what happens.

      Delete
    3. I'm not sure that is actually a great predictor.
      I got an 800 on SAT verbal meaning that my LSAT minimum score would be an impossible to receive 182…..

      Delete
    4. Same for me anon ^^ not expecting a 182 when my June score rolls in

      Delete
  5. Creepy.... One formula said I'd make a 158, another a 159, this one a 159 nearing 160. Then, using the comment prior to mine, I should have a 159. Well, that means I have to figure out how to get a higher score if I want to get into a top law school. Then again, I remember using maybe half my brain throughout my SAT practice classes and concentrating my actual studying into 10 hours the week before taking the actual test. :/

    I'm hoping that making a 4.0 so far in college is a good sign. ;)

    ReplyDelete
  6. I had a 1460 on my SAT with no studying whatsoever. Then again, I was 14 when I took it and am 26 now. Your formula says I should expect around a 171 on the LSAT but I think my intelligence may have decline a bit since then. I've also been out of school for a while. I wonder if, perhaps, your formula anticipates that those who did well on the SAT studied for it...perhaps study time is a larger factor in the LSAT than the SAT?

    ReplyDelete
  7. Attention everyone:

    A Pearson coefficient of R^2=0.3788 for LSAT / SAT correlation is quite weak. In words, that means that the LSAT / SAT relationship explains only 38% percent of the data points.

    Since the number of data points n=70, that means that this LSAT / SAT relationship only explains 27 (26.6 rounded) of the 70 data points. That's not very convincing.

    How about instead of spending this time worrying about our potential scores, we spend this time studying for the test?

    Best of luck, everyone! :)

    ReplyDelete
  8. I would be curious to know if these SAT scores are all from tests after 1996 because the SAT was rescaled then. So if you're mixing before and after 1996 it's meaningless.

    ReplyDelete