Causal Reasoning Flaws: The Six Ways LSAT Arguments Break

"Correlation does not imply causation" is the most repeated piece of advice in LSAT prep and the least useful, because it tells you nothing about what to look for. Every causal question asks the same thing — is this evidence enough to support the causal claim — and the answer is always no, for reasons that differ by type. Knowing the six patterns below is what lets you eliminate four of five choices without reading them.

Causal flaws are also unusually consistent in where they appear. They cluster in Strengthen, Weaken, and Flaw questions, and they show up whenever a stimulus describes two things changing together and then asserts that one produced the other.

The structure every causal argument has

Before the errors, the anatomy. A causal argument needs three things: a cause, an effect, and evidence linking them. The evidence is almost always an observed correlation — the two things vary together. The argument then asserts that the correlation establishes the direction, and that is where it breaks.

1. The claim

   Usually stated as "X causes Y" or "X leads to Y." Note whether the stimulus claims a universal relationship or a tendency, because a single counterexample destroys a universal claim and only weakens a tendency.

2. The evidence

   Frequently a correlation, but sometimes a plausible mechanism, an expert opinion, or an analogy. Each has its own weakness, and knowing which one is present tells you which fix the question wants.

3. The gap

   What the evidence does not establish. This is what Strengthen and Weaken target, and what a Flaw question asks you to name.

The six patterns

Each pattern has a distinct signature. Learning the signature is faster than reading every choice, because four of the five distractors will match a pattern you already know.

The first five are about evidence quality; the sixth is a translation error and is the easiest to spot.

1. Post hoc ergo propter hoc

One thing happens, then another happens, therefore the first caused the second. The stimulus almost always makes the sequence explicit and the timing suggestive. The fix is information showing the effect would have occurred anyway, or that the timing is coincidental.

2. Reverse causation

The direction of influence runs backwards, and often runs both ways. Sleep quality affects exam performance, and exam performance affects perceived sleep quality. The tell is that the effect is at least as plausible a cause of the apparent cause as the other way around — and in social science arguments it frequently is.

3. Common cause

Both variables are downstream of something unmentioned. Exercise correlates with lower mortality, but fitness may drive both exercise and health outcomes. The fix names the third variable, which makes the original evidence explainable without the causal claim.

4. Selection bias

The evidence came from a group that differs from the population being generalized about. This one appears frequently in stimulus about survey data or volunteer samples, and it is the pattern most often missed because the stimulus presents the data without visibly restricting it.

5. Coincidence

The correlation is real in the data but not systematic. The distinguishing feature is scale: a single observation or a tiny sample cannot establish a general claim. Weaken by showing the pattern does not replicate, or by offering a chance explanation.

6. Necessary condition mistaken for cause

A translation error rather than an evidence error. If X requires Y, then Y is necessary for X — but that does not make Y sufficient for X. This is the same structure as the conditional errors covered in our conditional reasoning guide, and it appears here whenever a stimulus uses "requires," "relies on," or "depends on."

What strengthens and weakens

Once you have identified the pattern, the answer choices become predictable because each one repairs a specific gap.

A wrong answer choice usually repairs a gap the argument does not have, or introduces a fourth variable unrelated to the mechanism.

A diagnostic you can use under pressure

1. Ask whether the evidence is correlation, mechanism, or analogy. This tells you which weakness is in play before you read the choices.
2. Ask whether the claim is universal ("always," "all") or a tendency. A single counterexample is decisive against the former and weak against the latter — and many answer choices exploit exactly that difference.
3. Ask whether the effect could produce the cause. If yes, the direction is unestablished regardless of how strong the correlation looks.
4. Only then read the choices. If you have named the pattern, you can often eliminate four without reading them fully.

Causal reasoning is one of the areas where untimed work pays off fastest, because the reasoning is slow and the answer choices are predictable. Drill it without the clock first, and keep a log of which pattern you missed — the same four patterns account for most of the misses, and they stop recurring once you can name them. For the broader question-type context, see our 14 Logical Reasoning question types guide.

Questions we get asked

LSAT® is a registered trademark of the Law School Admission Council, which does not review or endorse this site. This article is educational and is not legal or admissions advice. Format and policy figures reflect LSAC publications available as of October 2026 — confirm against lsat.lsac.org before you register.