LSAT is learnable: a theory Forum

[manwithplan]

One time I sat down and basically diagrammed a legal reasoning section of the LSAT with no timer. What I observed was that as one progressed through the section (that is, from question 1 to question 25 or 26), the percentage of the words in the prompt that one needed to actually incorporate in their thinking in order to get the question correctly was larger. In other words, what made the later questions "harder" was not so much that they were longer, or that the material itself was more complicated, but that a higher percentage of words within the prompt could be logically critical to the ideas. Basically, my theory is that what logical reasoning tests is your ability to keep track of an increasing number of factors that affects the logical answer to a question. In the early questions, the prompt can be very long, but ultimately, there are only going to be 1-3 little factors to keep track of. In the later questions, almost every word in the prompt creates a detail that can change the answer, so there are probably a dozens variables that could affect the answer. So basically, you just have to come up with an exercise that forces you to keep track of many factors which are practically disparate but logically connected, and do that over and over again. It makes sense if you think about it — they are showing you larger and larger amounts of data and making you sift through that data to find the correct answer.

There are two ways to beat this system: either you do the earlier, simpler problems in less time, leaving more time for the later problems, or you genuinely become better at the harder problems.

I don't think you have to get literally "smarter" in any fundamental way to do the harder problems better. All you need is an awareness that the test makers are looking to hinge the later questions on smaller and smaller little factors. So the key skill will be how quickly you can filter through the little factors.

Skills at the beginning of LR are totally different than skills at the end.

[Pneumonia]

Yes, the entire exam tests only a single skill. In games the premises are free. In LR they are free at the beginning and lightly disguised by the end. In reading comp they are heavily disguised and surrounded by irrelevancies.

[manwithplan]

Yes, the entire exam tests only a single skill. In games the premises are free. In LR they are free at the beginning and lightly disguised by the end. In reading comp they are heavily disguised and surrounded by irrelevancies.

In my next life, I will start (rather than finish) my LSAT journey with this insight, and finally break 175…

[liannabrianna]

One time I sat down and basically diagrammed a legal reasoning section of the LSAT with no timer. What I observed was that as one progressed through the section (that is, from question 1 to question 25 or 26), the percentage of the words in the prompt that one needed to actually incorporate in their thinking in order to get the question correctly was larger. In other words, what made the later questions "harder" was not so much that they were longer, or that the material itself was more complicated, but that a higher percentage of words within the prompt could be logically critical to the ideas. Basically, my theory is that what logical reasoning tests is your ability to keep track of an increasing number of factors that affects the logical answer to a question. In the early questions, the prompt can be very long, but ultimately, there are only going to be 1-3 little factors to keep track of. A problem solution essay can make any student's life easier. In the later questions, almost every word in the prompt creates a detail that can change the answer, so there are probably a dozens variables that could affect the answer. So basically, you just have to come up with an exercise that forces you to keep track of many factors which are practically disparate but logically connected, and do that over and over again. It makes sense if you think about it — they are showing you larger and larger amounts of data and making you sift through that data to find the correct answer.

There are two ways to beat this system: either you do the earlier, simpler problems in less time, leaving more time for the later problems, or you genuinely become better at the harder problems.

I don't think you have to get literally "smarter" in any fundamental way to do the harder problems better. All you need is an awareness that the test makers are looking to hinge the later questions on smaller and smaller little factors. So the key skill will be how quickly you can filter through the little factors.

Skills at the beginning of LR are totally different than skills at the end.

Uhmm, I bet that the difficulty of the last questions is not solely due to an increased number of factors. I mean, the complexity and length of the prompt, as well as the introduction of complex scenarios and nuanced arguments, contribute significantly to the heightened challenge. Additionally, I really believe that success on the LSAT requires a comprehensive set of skills throughout the entire test rather than distinct strategies for early and later questions...

[bajablast]

I went from a 146 to a 172 for an official score. It took me about a year to consistently score over 170 (but by the end I was hitting 175+ regularly on practice tests, and even scored a 180 once). It is definitely learnable, don't let anyone tell you otherwise.

[msft]

I went from a 146 to a 172 for an official score. It took me about a year to consistently score over 170 (but by the end I was hitting 175+ regularly on practice tests, and even scored a 180 once). It is definitely learnable, don't let anyone tell you otherwise.

agree 100%

I have seen average joes get near 180 just by practicing

It's actually quite amazing

had I known that 40 years ago things may have been different for me

[manwithplan]

One time I sat down and basically diagrammed a legal reasoning section of the LSAT with no timer. What I observed was that as one progressed through the section (that is, from question 1 to question 25 or 26), the percentage of the words in the prompt that one needed to actually incorporate in their thinking in order to get the question correctly was larger. In other words, what made the later questions "harder" was not so much that they were longer, or that the material itself was more complicated, but that a higher percentage of words within the prompt could be logically critical to the ideas. Basically, my theory is that what logical reasoning tests is your ability to keep track of an increasing number of factors that affects the logical answer to a question. In the early questions, the prompt can be very long, but ultimately, there are only going to be 1-3 little factors to keep track of. A problem solution essay can make any student's life easier. In the later questions, almost every word in the prompt creates a detail that can change the answer, so there are probably a dozens variables that could affect the answer. So basically, you just have to come up with an exercise that forces you to keep track of many factors which are practically disparate but logically connected, and do that over and over again. It makes sense if you think about it — they are showing you larger and larger amounts of data and making you sift through that data to find the correct answer.

There are two ways to beat this system: either you do the earlier, simpler problems in less time, leaving more time for the later problems, or you genuinely become better at the harder problems.

I don't think you have to get literally "smarter" in any fundamental way to do the harder problems better. All you need is an awareness that the test makers are looking to hinge the later questions on smaller and smaller little factors. So the key skill will be how quickly you can filter through the little factors.

Skills at the beginning of LR are totally different than skills at the end.

Uhmm, I bet that the difficulty of the last questions is not solely due to an increased number of factors. I mean, the complexity and length of the prompt, as well as the introduction of complex scenarios and nuanced arguments, contribute significantly to the heightened challenge. Additionally, I really believe that success on the LSAT requires a comprehensive set of skills throughout the entire test rather than distinct strategies for early and later questions...

"Complex scenarios" is just another way of saying "needing to keep track of more premises,"
which is my entire point.

[Marty Murray]

I went from a 146 to a 172 for an official score. It took me about a year to consistently score over 170 (but by the end I was hitting 175+ regularly on practice tests, and even scored a 180 once). It is definitely learnable, don't let anyone tell you otherwise.

I totally agree, and even little improvements such as identifying conclusions more exactly or reading answer choices more carefully can result in big score increases.

[nealric]

Standardized tests in general seek to evaluate your ability to perform a particular cognitive task (or set of tasks). They don't always succeed in this- but that's a different story.

If you take a standardized test cold, you are using some amount of time and mental effort simply to understand what is being asked of you. So there's low-hanging fruit of improvement in just basic familiarization with the test.

As you practice further, you can build a memorized catalog of the type of questions that may be asked. For example, there are only so many variations in logic game types. Once you know all the potential types of games and have identified a strategy for each type, you again greatly reduce the time and mental effort required to solve a given question. In a standardized test with tight time constraints, efficiency is huge. I suspect the vast majority of sub-160 scorers either do not complete each section or decide to guess before having completely analyzed some questions in the interest of time.

Think of it like a basic arithmetic problem like long division. People with very high natural aptitude for arithmetic might immediately intuit the algorithm for performing long division. But most people when first exposed to a long division problems would either take a very long time to arrive at the answer or be unable solve it. However, most children of average aptitude are able to learn long division by the age of 9 or 10 by having the algorithm explained and practicing its use extensively. It's a similar story with most LSAT question types.

As a side note, I mention logic games because they are probably the most learnable in the context of studying for one specific test. Most people have spent a lot of time in their academic past doing reasoning similar to what is asked in the reading comp and logical reasoning sections, but games is often new. That means, there's often a lot more room for improvement (not that other sections can't be improved upon).