Why Your LSAT Reading Comp Dropped While Your LR Improved

If your Logical Reasoning score climbed and your Reading Comprehension score slipped over the same stretch, the two are probably related. The reading posture that Logical Reasoning trains — skeptical, argument-hunting, alert to every gap — is not the posture Reading Comprehension rewards.

This surprises people because both sections involve reading, so improvement in one seems like it should transfer to the other. But the sections ask for different things. Logical Reasoning asks you to evaluate a short argument critically. Reading Comprehension mostly asks you to understand a long passage accurately and then locate evidence in it.

Understanding and evaluating are different mental operations, and getting much better at one of them can crowd out the other. That is the mechanism behind the pattern, and it is fixable once you can see it.

The useful shorthand is that Logical Reasoning wants skepticism and Reading Comprehension wants curiosity. In LR you are looking for what is wrong with an argument. In RC you are trying to follow what someone is saying.

The last row is the one that decides scores. Reading Comprehension is closer to an evidence-finding task than an analysis task. The correct answer is generally supported by something the passage actually says, and your job is to locate that something rather than to reason your way toward the most defensible-sounding claim.

There is genuine overlap at the edges — RC includes a few questions that ask you to strengthen or weaken something, and occasional passages turn on fine distinctions between words. But the overlap is small enough that the sections should be trained separately.

The first is arguing with the passage. In LR, noticing that a claim is under-supported is the whole job. In RC, an author can assert something thinly and it does not matter — you are not being asked whether they are right. Students who have trained skepticism hard start mentally objecting to passages, and every objection is attention spent on a question that will not be asked.

The second is over-analyzing sentence by sentence. Logical Reasoning rewards slowing down on a five-line stimulus and extracting exactly what each sentence commits the author to. Applied to a sixty-line passage, that same care produces a detailed understanding of individual sentences and no grasp of how the passage is organized — which is what most of the questions are about. You finish the passage having understood everything and retained nothing structural.

The third is over-notating. Diagramming and heavy annotation help in LR precisely because the text is short and dense. In RC, extensive marking often becomes a substitute for comprehension: you produce a heavily highlighted passage and a vague memory of it. The purpose of a mark is to help you find something again, not to prove you read carefully.

All three habits share a signature: they make you slower and more thorough, and in RC that combination frequently makes you worse.

Reading Comprehension scores tend to be less stable than Logical Reasoning scores, and the most common driver is not skill. It is topic.

The questions do not really change from section to section — the same handful of question types recur. What changes is subject matter. One section hands you a passage about a legal doctrine and one about a musical movement; the next hands you something about mineral chemistry. If one of those topics loses you, comprehension degrades early, and every question attached to that passage gets harder at once.

That is why RC results often look erratic in a way LR results do not. A bad RC section is frequently one bad passage, and one bad passage can account for a quarter of the section.

The practical implication is that manufactured interest is a real technique rather than a soft one. Approaching a passage as something you are going to have to explain to a friend afterward changes how much of it you retain, and it costs nothing. Students who find this hard to take seriously usually have not noticed how much of their RC variance sits in the passages they found boring.

The mistake underneath all of this is applying LR's method at RC's scale. It feels responsible — you are reading more carefully, thinking harder about each claim, being appropriately critical. It is also the thing making the section worse.

There is a recognizable version of this student: strong analytical background, LR score climbing steadily, RC stuck or sliding, and completely certain the problem is reading speed. They respond by trying to read faster, which strips out comprehension while leaving the over-analysis intact, and the score moves the wrong way again.

Genuine reading-speed problems exist but are much rarer than they are diagnosed. Far more often, the slowness is a symptom: you are reading slowly because you are analyzing every sentence, and you are analyzing every sentence because that is the habit the other section built.

The second version of the mistake is deciding RC is a fixed trait — that some people are just good at it. Reading Comprehension responds to method at least as much as Logical Reasoning does. It just responds to a different method.

Split the work in two. On the first read, get the structure. On the questions, get the evidence.

Structural reading means tracking what the passage is doing rather than memorizing what it says. Who holds which view, where the author's own voice enters, what each paragraph contributes, where the passage turns. You do not need to retain the details on the first pass — details are still on the page and you can go get them. Structure is what tells you where to go get them.

Then, on every question, hold yourself to finding the specific piece of text that supports your answer. If you cannot point to a sentence or phrase that makes the answer true, you are probably on the wrong answer. This single rule handles a large share of RC misses, because most tempting wrong answers are ones that sound consistent with the passage's general drift while having nothing behind them.

That evidence requirement is also the cleanest way to break the arguing habit. It gives your skepticism somewhere legitimate to go: instead of interrogating the author, interrogate your own answer choice.

Suppose a passage discusses a historian who argues that a city's rapid growth in the 1800s was driven mainly by a new canal, and then presents a later scholar who contends that regional migration patterns already underway explain most of the growth. The author closes by noting that the second account better fits the timing of the population data, while conceding the canal likely accelerated an existing trend.

In LR mode, you would start evaluating. Is population timing really sufficient to establish that? Has the second scholar ruled out the canal as a partial cause? Those are reasonable thoughts and none of them will be asked. Meanwhile you may not have registered the thing the questions will turn on: the author does not fully endorse either scholar and lands on a qualified position.

In RC mode, the passage reads as a small map. Two competing explanations, a body of evidence that favors the second, and an author who leans that way while granting the first some role. If a question asks for the author's attitude and one choice says the author rejects the canal explanation, you can eliminate it immediately by pointing at the concession in the last sentence.

Same passage, same amount of reading. The difference is what you were reading for.

Practice the two sections in separate blocks rather than mixing them within a study session, at least while you are rebuilding the RC posture. You are trying to make two distinct modes feel distinct, and alternating between them every few minutes works against that.

Before each RC passage, take a deliberate second to change gears. Something as simple as telling yourself you are about to learn what this person thinks, not judge it, is enough to interrupt the default. It sounds trivial and it measurably changes what you notice.

In review, add one question to your RC process: for each miss, could you have pointed to the sentence supporting the right answer? If yes, the failure was answer selection. If no, the failure was comprehension, and the fix is in how you read the passage rather than how you evaluated the choices.

Finally, get several passages fully right before worrying about finishing all four. Accuracy on three passages you actually understood is a stronger foundation than partial coverage of all four, and the speed tends to follow the comprehension rather than the other way around.