Do people who over study for the LSAT do worse in law school Forum

[jk148706]

Related q:


do ppl who study and game law school exams actually do worse at firms? (!!!!)

[Straw_Mandible]

I think the problem with this reasoning is that when a person hasn't prepped, it is safer to assume that they are scoring well because of a "natural ability in the specific skills tested by the LSAT" -- i.e., if they do well in the logical reasoning, it is probably because they are good at logical reasoning; if they are good at the reading comprehension, that's probably a skill they have in day-to-day life. We aren't assuming they possess higher cognitive abilities in skills not tested by the LSAT, but rather that their score is indicative of ability in the skills that the LSAT tries to approximate.

With the person who has prepped, there is a confounding factor. We don't know if that person is doing better because they have learned the actual skill that the test is trying to approximate, or because they have learned to perform the task required by the test.
The LSAT does not directly test the skills it tests, it approximates them. A logic game is a way of approximating logical reasoning. But you can improve your ability of solving a LG in many ways/strategies that probably don't do much for your day-to-day logical reasoning skills. Sure, it might help you to think other things through more logically, but it is possible that you have developed a tool (i.e. a diagram), but have not truly expanded the actual logical reasoning capacity.

This is an interesting thought. But I'm not sure that I totally understand what you mean when you say that the LSAT tests skills "indirectly."

It's entirely obvious to me that anyone (preparation or no preparation) who scores -0 on a RC section is a skilled reader to the extent that reading skills are measured by the LSAT. It's equally obvious that anyone who scores -0 on a LG section possesses reasoning skills to the extent that reasoning skills are measured by the LSAT.

So to assume that someone who started out at -0 in RC has "real" reading skills, while someone who devoted a year of her life to learning how to go -0 in RC (i.e., learning exactly those reading skills which are tested by the LSAT) somehow has "fake" reading skills seems like a strange thing to say.

Is it not equally possible that someone who starts out at -0 in RC possesses exactly those reading skills which are tested by the LSAT (e.g. the ability to quickly grasp a big-picture understanding of reasoning structure while glossing over specific content and details), but lacks certain reading skills that may be advantageous in other contexts (such as the ability to read slowly, deeply, and closely and to make connections across a range of subject matter based on certain specific details)? My point is that it's not enough to say that this person is a "good" reader, just like it's not enough to say that this person is "intelligent."

The only thing we can extrapolate from the fact that a person scored in the 99th percentile on the LSAT is that this person has the skills necessary to score in the 99th percentile on the LSAT. And that statement applies equally to people who achieve that score with little or no prep, and people who spend a year preparing in order to achieve that score.

[09042014]

I think the problem with this reasoning is that when a person hasn't prepped, it is safer to assume that they are scoring well because of a "natural ability in the specific skills tested by the LSAT" -- i.e., if they do well in the logical reasoning, it is probably because they are good at logical reasoning; if they are good at the reading comprehension, that's probably a skill they have in day-to-day life. We aren't assuming they possess higher cognitive abilities in skills not tested by the LSAT, but rather that their score is indicative of ability in the skills that the LSAT tries to approximate.

With the person who has prepped, there is a confounding factor. We don't know if that person is doing better because they have learned the actual skill that the test is trying to approximate, or because they have learned to perform the task required by the test.
The LSAT does not directly test the skills it tests, it approximates them. A logic game is a way of approximating logical reasoning. But you can improve your ability of solving a LG in many ways/strategies that probably don't do much for your day-to-day logical reasoning skills. Sure, it might help you to think other things through more logically, but it is possible that you have developed a tool (i.e. a diagram), but have not truly expanded the actual logical reasoning capacity.

This is an interesting thought. But I'm not sure that I totally understand what you mean when you say that the LSAT tests skills "indirectly."

It's entirely obvious to me that anyone (preparation or no preparation) who scores -0 on a RC section is a skilled reader to the extent that reading skills are measured by the LSAT. It's equally obvious that anyone who scores -0 on a LG section possesses reasoning skills to the extent that reasoning skills are measured by the LSAT.

So to assume that someone who started out at -0 in RC has "real" reading skills, while someone who devoted a year of her life to learning how to go -0 in RC (i.e., learning exactly those reading skills which are tested by the LSAT) somehow has "fake" reading skills seems like a strange thing to say.

Is it not equally possible that someone who starts out at -0 in RC possesses exactly those reading skills which are tested by the LSAT (e.g. the ability to quickly grasp a big-picture understanding of reasoning structure while glossing over specific content and details), but lacks certain reading skills that may be advantageous in other contexts (such as the ability to read slowly, deeply, and closely and to make connections across a range of subject matter based on certain specific details)? My point is that it's not enough to say that this person is a "good" reader, just like it's not enough to say that this person is "intelligent."

The only thing we can extrapolate from the fact that a person scored in the 99th percentile on the LSAT is that this person has the skills necessary to score in the 99th percentile on the LSAT. And that statement applies equally to people who achieve that score with little or no prep, and people who spend a year preparing in order to achieve that score.

Do you understand our arguments or not? Because it doesn't seem like you do.

[rpupkin]

Do you understand our arguments or not? Because it doesn't seem like you do.

The arguments are complex and cannot be comprehended without expert logical reasoning skills. In other words, he won't be able to understand the arguments without first doing some more LSAT studying.

[Straw_Mandible]

Do you understand our arguments or not? Because it doesn't seem like you do.

I do. You're arguing that a person who prepares for the LSAT is learning how to pretend that they have reading and reasoning skills. I am arguing that they are learning exactly those reading and reasoning skills which are necessary to perform highly on the LSAT.

I've made my claim obvious. Yours still needs some support.

[Theopliske8711]

You're arguing that a person who prepares for the LSAT is learning how to pretend that they have reading and reasoning skills.

That's not what we're arguing. At least not I.

[beepboopbeep]

Do you understand our arguments or not? Because it doesn't seem like you do.

I do. You're arguing that a person who prepares for the LSAT is learning how to pretend that they have reading and reasoning skills. I am arguing that they are learning exactly those reading and reasoning skills which are necessary to perform highly on the LSAT.

I've made my claim obvious. Yours still needs some support.

Appropriate username.

[09042014]

Do you understand our arguments or not? Because it doesn't seem like you do.

I do. You're arguing that a person who prepares for the LSAT is learning how to pretend that they have reading and reasoning skills. I am arguing that they are learning exactly those reading and reasoning skills which are necessary to perform highly on the LSAT.

I've made my claim obvious. Yours still needs some support.

Not that's it's pretend, but that they learn strategies that only apply to LSAT and not general reasoning ability. And by doing so it ruins the LSAT's predictive ability.

I'm going to use the Soduku example. If you take someone who does one every day, and compare them against somoene who did 2-3 in their life, and they scored the same. Would say they are just as good as analytically reasoning? Or would you say the first person just practiced Soduku a lot? If you switched to a new game do you think they'd still tie?

[Theopliske8711]

I think Straw would realize what we are saying if he realized that the LSAT is a man-made test, and therefore is pretty damn imperfect. It tries to test certain skills, but, like so many man-made things, it has serious flaws that can undermine it's goal. The point of the testprep is to take advantage of that flaw.

[Straw_Mandible]

Not that's it's pretend, but that they learn strategies that only apply to LSAT and not general reasoning ability. And by doing so it ruins the LSAT's predictive ability.

I'm going to use the Soduku example. If you take someone who does one every day, and compare them against somoene who did 2-3 in their life, and they scored the same. Would say they are just as good as analytically reasoning? Or would you say the first person just practiced Soduku a lot? If you switched to a new game do you think they'd still tie?

Right. And I'm questioning the idea of "general reasoning ability" as a meaningful metric.

The Sudoku example is a good one. It seems intuitively true that the person who caught on to Sudoku very quickly should also be able to solve other logic puzzles very quickly. But my point is that we just don't know how each person's cognitive skills will align with a different task. We also don't know how much the first person improved his reasoning skills in the process of practicing Sudoku. The only thing we know is that they are equally good at Sudoku.

Edit: I know that the argument I'm making feels strange and counter-intuitive. That could be because it's wrong. On the other hand, it could be because it goes against the way we customarily understand intelligence, which may itself be wrong. But people ITT are throwing around a lot of assumptions about "intellect" and "general reasoning ability," and I think we should do our best to bring those assumptions out into the open.

[quijotesca1011]

Not that's it's pretend, but that they learn strategies that only apply to LSAT and not general reasoning ability. And by doing so it ruins the LSAT's predictive ability.

I'm going to use the Soduku example. If you take someone who does one every day, and compare them against somoene who did 2-3 in their life, and they scored the same. Would say they are just as good as analytically reasoning? Or would you say the first person just practiced Soduku a lot? If you switched to a new game do you think they'd still tie?

Right. And I'm questioning the idea of "general reasoning ability" as a meaningful metric.

The Sudoku example is a good one. It seems intuitively true that the person who caught on to Sudoku very quickly should also be able to solve other logic puzzles very quickly. But my point is that we just don't know how each person's cognitive skills will align with a different task. We also don't know how much the first person improved his reasoning skills in the process of practicing Sudoku. The only thing we know is that they are equally good at Sudoku.

Essentially, it's a better bet that the person who scored high tested cold doesn't just have strong cognitive abilities in the one area being tested. The test is like a snapshot that captures a particular piece of what is most likely a larger well of logical reasoning ability. Is that true for everyone? Maybe not. But it's a pretty good bet that someone who scores 99 percentile on the LSAT with no preparation has good logical reasoning skills and would perform well on other assessments. And it is hard to imagine someone being naturally 'good' at only the LSAT questions and not at other logical reasoning tasks.

In contrast, someone who has studied to the LSAT has learned the skills necessary to master that particular logical reasoning assessment. I'm not saying that they are pretending, but they have learned a very particular skill set within logical reasoning skills. That is different, in my mind, from having a high natural base of logical reasoning, and the suggestion that by studying for the LSAT they now have an equivalent base level of broad-based logical reasoning skills, to me is tenuous. If I had to bet on one of the two sudoku-ers taking on a new cognitive task, I would go with the one who only did it 2-3 times. There is a higher probability that that person's overall, free-standing logical reasoning / cognitive skills are naturally stronger and would adapt more easily to a new test. With the person who did sudokus daily, I am sure they could reach the same strength on the second test. I am just not sure that they could do so as quickly as the other person - they might very well have to do significant preparation for that test as well.

[lawhopeful10]

This is just anecdotal but I have had the same thoughts DF has before. I slacked off in college and thus didn't have as great a GPA or LSAT as I thought I could. Then I did really well on my law exams because I actually did work hard in law school so my potential was much higher then people who had basically already maxed out via working much harder earlier. As others have said though there is no way for me to prove this it's just my thought.

[$alty]

what do you think the avg study time is for a 170+ scorer? i know lots who took classes or studied a ton on their own. could it be so much that there are so few people who put in no prep to reach that level and therefore it doesn't even matter?

[09042014]

what do you think the avg study time is for a 170+ scorer? i know lots who took classes or studied a ton on their own. could it be so much that there are so few people who put in no prep to reach that level and therefore it doesn't even matter?

It would be interesting to have some data on that. I'd bet the average scorer did some prep. But the amount talked about on TLS is way above and beyond.

[d cooper]

Gunning the LSAT won't make you worse, it'll just give you an LSAT score above your actual worth, meaning you will under perform in law school because you went to a better law school than you should have. Of course this on average, an individual LSAT doesn't mean shit.

This assumes you're competing with a bunch of naturals.

Since grades are assigned relative to your peers, a relevant consideration is the percentage of the class that studied for their LSAT score. I would suspect that most T14 students these days are putting in a lot of study time just because the materials are available and these students generally have a track record of over-achieving. If there are just a handful of natural 175+ scorers in your class, and even assuming that that makes them objectively better at law school, that won't substantially impact your performance in school with regard to rank.

But even if the LSAT naturals are cognitively gifted and that somehow translates to better law school grades, it still makes sense to game the test as much as possible so you can end up in a pool of students more likely to get good jobs. No one should be suggesting that all of the H-bound 17X self-study applicants should have just gone to a TTT with their 162 so that they could be in a pool of students that equal their "actual worth."

[beepboopbeep]

Not that's it's pretend, but that they learn strategies that only apply to LSAT and not general reasoning ability. And by doing so it ruins the LSAT's predictive ability.

I'm going to use the Soduku example. If you take someone who does one every day, and compare them against somoene who did 2-3 in their life, and they scored the same. Would say they are just as good as analytically reasoning? Or would you say the first person just practiced Soduku a lot? If you switched to a new game do you think they'd still tie?

Right. And I'm questioning the idea of "general reasoning ability" as a meaningful metric.

The Sudoku example is a good one. It seems intuitively true that the person who caught on to Sudoku very quickly should also be able to solve other logic puzzles very quickly. But my point is that we just don't know how each person's cognitive skills will align with a different task. We also don't know how much the first person improved his reasoning skills in the process of practicing Sudoku. The only thing we know is that they are equally good at Sudoku.

Essentially, it's a better bet that the person who scored high tested cold doesn't just have strong cognitive abilities in the one area being tested. The test is like a snapshot that captures a particular piece of what is most likely a larger well of logical reasoning ability. Is that true for everyone? Maybe not. But it's a pretty good bet that someone who scores 99 percentile on the LSAT with no preparation has good logical reasoning skills and would perform well on other assessments. And it is hard to imagine someone being naturally 'good' at only the LSAT questions and not at other logical reasoning tasks.

In contrast, someone who has studied to the LSAT has learned the skills necessary to master that particular logical reasoning assessment. I'm not saying that they are pretending, but they have learned a very particular skill set within logical reasoning skills. That is different, in my mind, from having a high natural base of logical reasoning, and the suggestion that by studying for the LSAT they now have an equivalent base level of broad-based logical reasoning skills, to me is tenuous. If I had to bet on one of the two sudoku-ers taking on a new cognitive task, I would go with the one who only did it 2-3 times. There is a higher probability that that person's overall, free-standing logical reasoning / cognitive skills are naturally stronger and would adapt more easily to a new test. With the person who did sudokus daily, I am sure they could reach the same strength on the second test. I am just not sure that they could do so as quickly as the other person - they might very well have to do significant preparation for that test as well.

I agree with the basic premise, but this is where it gets more complicated - while I don't agree with Straw Man overall, I'm also not sure that there'd be a huge correlation between low amounts of LSAT studying and high law school grades, ceteris paribus. The person doing the prep work is learning how to take a test, and that's an important skill for lawl school. They may be better at identifying weaknesses and filling in gaps in their test ability than the natural. Then again, they may be more likely to view law school exams as similarly mechanical to the LSAT, and not as quick to intuit how to tailor their answers to a particular professor's whims, or less able to make creative connections between ideas on the spot. Lots of assumptions on the table at this point.

The sudoku comparison would be more apt if we threw people into law school exams without prep. But of course that's not how it works.

[09042014]

Gunning the LSAT won't make you worse, it'll just give you an LSAT score above your actual worth, meaning you will under perform in law school because you went to a better law school than you should have. Of course this on average, an individual LSAT doesn't mean shit.

This assumes you're competing with a bunch of naturals.

Since grades are assigned relative to your peers, a relevant consideration is the percentage of the class that studied for their LSAT score. I would suspect that most T14 students these days are putting in a lot of study time just because the materials are available and these students generally have a track record of over-achieving. If there are just a handful of natural 175+ scorers in your class, and even assuming that that makes them objectively better at law school, that won't substantially impact your performance in school with regard to rank.

But even if the LSAT naturals are cognitively gifted and that somehow translates to better law school grades, it still makes sense to game the test as much as possible so you can end up in a pool of students more likely to get good jobs. No one should be suggesting that all of the H-bound 17X self-study applicants should have just gone to a TTT with their 162 so that they could be in a pool of students that equal their "actual worth."

You are right that I bet most people in the t14 aren't taking it cold, but they probably aren't self studying nearly the degree that TLS does. If they did, I believe you'd see a significant rise the percentiles associated with LSAT percentiles. Though I haven't been paying attention, maybe that has changed.

It definitely makes sense to game it as hard as you can.

[Straw_Mandible]

One last point: Many people in this thread have suggested that when people super-study for the LSAT, they are focusing their attention on learning "tricks" and "strategies," which allow them to out-perform their actual skill levels. That's really not how any of us (the successful ones, at least) are approaching this. We assess and target our weaknesses in terms of the general skills we need to improve upon, and then we devote hours upon hours of effort to focused, deliberate practice of those skills. It has nothing to do with "gaming" the test. When I drill RC sections, all of my mental energy goes into strengthening my ability to understand dense and convoluted phrases and sentences as quickly as possible, identify main points, organize details concisely and clearly, and see the rational progression of an argument or series of arguments. These are all very generalized reading skills, which are necessary for improvement in the context of the LSAT, but they are also extremely useful in almost every other context of reading that I can think of. As far as I know, there is no way to meaningfully increase your score without actually learning these skills. But maybe I'm doing it wrong.

Now, if what you're saying is true--and there is a way to improve at LSAT reading comp without actually becoming a better reader--someone please tell us what that is, so that we can all stop wasting away our precious nights and weekends in the library.

[180kickflip]

One last point: Many people in this thread have suggested that when people super-study for the LSAT, they are focusing their attention on learning "tricks" and "strategies," which allow them to out-perform their actual skill levels. That's really not how any of us (the successful ones, at least) are approaching this. We assess and target our weaknesses in terms of the general skills we need to improve upon. It has nothing to do with "gaming" the test. When I drill RC sections, all of my mental energy goes into strengthening my ability to understand dense and convoluted phrases and sentences as quickly as possible, identify main points, organize details concisely and clearly, and see the rational progression of an argument or series of arguments. These are all very generalized reading skills, which are necessary for improvement in the context of the LSAT, but they are also extremely useful in almost every other context of reading that I can think of. As far as I know, there is no way to meaningfully increase your score without actually learning these skills. But maybe I'm doing it wrong.

Now, if what you're saying is true--and there is a way to improve at LSAT reading comp without actually becoming a better reader--someone please tell us what that is, so that we can all stop wasting away our precious nights and weekends in the library.

I think focusing only on reading comp misses a bit of the issue. Reading comp is often mentioned as the most difficult section to improve on specifically because there aren't as many strategies and tricks to beating it. There are for LR and logic games. I think it's probably pretty rare to find someone who improved drastically on LG and did not use any tricks or strategies that were LSAT specific.

[Pneumonia]

It doesn't take away from your effort that 95% of what you're doing (including RC prep) is LSAT-specific, but it is still true. I think you're way overestimating the degree to which you're becoming a better reader and way underestimating the degree to which you're just recognizing patterns. Yes there is overlap, but not very much. Otherwise the only way to improve RC would be to retake life and read lots of books. The reason the best way to improve LSAT RC is by studying LSAT RC is that it is its own skill set.

I do agree that you will pick up some transferrable skills along the way, but not nearly enough to bridge the difference between natural and studier.

[Balthy]

It doesn't take away from your effort that 95% of what you're doing (including RC prep) is LSAT-specific, but it is still true. I think you're way overestimating the degree to which you're becoming a better reader and way underestimating the degree to which you're just recognizing patterns. Yes there is overlap, but not very much. Otherwise the only way to improve RC would be to retake life and read lots of books. The reason the best way to improve LSAT RC is by studying LSAT RC is that it is its own skill set.

I do agree that you will pick up some transferrable skills along the way, but not nearly enough to bridge the difference between natural and studier.

But those patterns you're learning to recognize are not lsat-specific. If you make a list of all the logical fallacies or types of arguments you'll hear throughout your life, it wouldn't be a huge list.

I do think it's possible to just learn "tricks" but it seems most prep companies have moved to encouraging a thorough understanding of the logic behind the questions you'll encounter.

[Pneumonia]

Discounting the issue of retention I will just reiterate that I think y'all are waaay overestimating how much of any prep material is substantive vs. situational. Again there is substantive stuff included, and therefore arguably retained, but it is a small percentage. I have taught, tutored and written (some) curriculum for a prep company btw (think PS/BP/Manhattan; ie not a shit tier company) so I'm not just speculating (not that anyone else is either).

There is legit room for disagreement, but my experience with people after the fact just does not at all bear out for me that anyone is materially improving any real skills.

[Straw_Mandible]

It doesn't take away from your effort that 95% of what you're doing (including RC prep) is LSAT-specific, but it is still true. I think you're way overestimating the degree to which you're becoming a better reader and way underestimating the degree to which you're just recognizing patterns. Yes there is overlap, but not very much. Otherwise the only way to improve RC would be to retake life and read lots of books. The reason the best way to improve LSAT RC is by studying LSAT RC is that it is its own skill set.

I do agree that you will pick up some transferrable skills along the way, but not nearly enough to bridge the difference between natural and studier.

But those patterns you're learning to recognize are not lsat-specific. If you make a list of all the logical fallacies or types of arguments you'll hear throughout your life, it wouldn't be a huge list.

I do think it's possible to just learn "tricks" but it seems most prep companies have moved to encouraging a thorough understanding of the logic behind the questions you'll encounter.

Exactly. And this is because the good prep companies realize that actually understanding that logic is the only way to move from the 150s into the 170s. You can't make that kind of jump just by remembering that any answer choice with the word "some" is always incorrect--or something absurd like that.

I made this point before, but even with the obvious "strategies" like diagramming in LG, it's impossible to properly apply those diagramming "tricks" unless you actually have a complete, intuitive understanding of the reasoning relationships tested by that game. Diagrams just enable you to organize the information you're given. They don't make logical inferences for you, and they don't solve questions for you. For that, you need to use real reasoning skills.

[Balthy]

I'm curious, is speed a big issue in law school exams? I think most people who go from 160s to 70s probably already understand all or almost all of the logical issues found in the LSAT, and their prep and progress have mostly to do with speed (which, of course, is somewhat related to better understanding). This is why the most important part of prep for most ppl on tls seems to be drilling. If this debate is really about natural speed vs trained speed and law school exams aren't as time-sensitive, then obviously the trained person shouldn't be at a disadvantage.

Also, Pneumonia, would you say the same thing about only those of your students who went up to the 170s? I'm not 100% confident on this issue but with people who get to 170s i tend to agree with mandible.

[A. Nony Mouse]

Depends on the prof and the exam, but many law school exams are very time sensitive - the person who can write down more points/spot more issues in 3 hours wins.