LSAT Parallel Reasoning

Parallel reasoning questions present an argument and ask you to choose another argument with similar reasoning. Parallel-flaw questions add that the original argument contains a flaw, so the answer must match the flaw as well as the structure. These are LSAT Logical Reasoning tasks: the key is form, not shared vocabulary or topic.

Identify the argument's skeleton

Before reading choices, reduce the stimulus to a short pattern. Mark the premises, conclusion, quantifiers, conditional arrows, and any causal or comparative move. You might summarize it as “all A are B; this is B; therefore it is A,” or “outcome rose after policy; policy caused outcome.” The summary should capture the logic and the flaw, not restate the story.

Then read the stem carefully. A parallel-reasoning question asks for similar reasoning, while a parallel-flaw question asks for an argument with the same reasoning error. A valid argument can be parallel to another valid argument. A flawed original requires a flawed match only when the stem specifies that the reasoning is flawed or asks for the same flaw.

Subject matter is often deliberately different. An argument about a school schedule can parallel one about a garden, a machine, or a company. If a choice repeats many words from the stimulus but changes the logical direction, it is not parallel. If it uses different vocabulary but the same structure, it may be the match.

Match conditional direction

Conditional arguments are common in parallel questions because a reversed arrow can be hard to notice in prose. If the stimulus says A → B, then concludes A from B, the flaw is a converse error. A parallel choice must also infer A from B without another premise establishing the reverse.

Original stimulus: “Every sculpture selected for the exhibition receives a catalog entry. This sculpture has a catalog entry. So it was selected for the exhibition.” Pattern: selected → entry; entry; therefore selected. A parallel flaw is: “Every permit issued for a street fair has a safety review. This event had a safety review. Therefore, it received a permit.” The review may be required for permits but may also occur for other events. Both arguments reverse a sufficient condition.

A choice that says “Every catalog entry belongs to a selected sculpture” would not match, because it supplies the reverse relationship. A choice concluding that the sculpture has an entry from its selection could be a valid argument, but it would not reproduce the original flaw. Track the arrow and the conclusion, not just the terms.

Match quantifiers and scope

Quantifier changes alter the argument. “Some A are B” is weaker than “all A are B,” and “most A are B” does not identify a particular member. An answer choice can resemble the original but use “all” where it said “some,” or generalize from one case to a whole group. That may change the structure or introduce a different flaw.

Suppose the stimulus says “Some trained volunteers can repair the device. Ren is a trained volunteer. Therefore, Ren can repair the device.” The premise about some volunteers does not establish that Ren is among them. The flaw is moving from existence in a group to an individual. A parallel answer should make the same unsupported attribution, not merely contain another “some” statement.

Scope also includes time, place, and population. If the original uses a result from one city to recommend a policy everywhere, a parallel argument should make a similar geographic leap. An answer that compares two similar cities or supplies evidence of representativeness may be valid or stronger, so it will not match the same unsupported generalization.

Match causal and analogical reasoning

For a causal argument, identify whether the stimulus moves from sequence or correlation to cause, ignores alternatives, or infers the wrong causal direction. A parallel answer should make the same kind of causal move. Merely reporting two events that happened together is not enough if the conclusion does not claim a causal relationship.

For an analogy, identify the similarities used to support a conclusion and ask whether the examples differ in a relevant way. A parallel choice should rely on a comparable analogy pattern. An answer using a conditional reversal may also be flawed, but it does not parallel an analogy flaw simply because both arguments are invalid.

Example: “A small town's visitor center increased tourism after adding a local-art exhibit. A nearby county should do the same because its residents also make art.” The conclusion relies on a similarity, but the evidence does not show that local art was the cause of the town's tourism increase or that the county has similar visitor demand. A parallel flaw could make a recommendation based on a superficial similarity while ignoring the difference that matters to the outcome.

A full worked question

Stimulus: “Every concert that sold out had an online presale. The orchestra's concert had an online presale, so it must have sold out.” The form is sold out → presale; presale; therefore sold out. The argument mistakes a necessary condition for a sufficient one, because a presale may occur even when seats remain.

Choice A: “Every student who completed the lab earned a certificate. Mina earned a certificate, so she completed the lab.” Same form, same converse error. Choice B: “Every sold-out concert had a presale. The orchestra concert sold out, so it had a presale.” Same topic but valid direction; not parallel. Choice C: “Some concerts had presales, and some sold out; therefore every presale sold out.” This commits a broad quantifier error, not the same conditional reversal. Choice D: “The orchestra sold out its concert after lowering ticket prices, so the lower prices caused the sale.” This is a causal inference. Choice A matches.

Choice A is the best match even though it changes concerts to students. The decisive pattern is “all outcomes have feature; this case has feature; therefore outcome.” The context is incidental. If the stem instead asked for a valid parallel argument, choice B would be the relevant structure because it uses the sufficient condition in the correct direction.

Parallel reasoning versus analogy questions

Parallel reasoning is about matching an argument's form. An analogy question asks whether similarities between cases support a conclusion. The two can overlap: a parallel argument can itself be analogical. But do not assume every parallel question calls for finding similar real-world cases. Compare the inference structure first.

A useful way to distinguish the task is to read the stem. If it asks which argument is most similar in its reasoning, match the structure. If it asks what follows or what assumption an analogy requires, evaluate the analogy itself. The stem sets the standard.

Make the comparison efficiently

Parallel choices can be long. Avoid analyzing every detail equally. First eliminate choices with the wrong conclusion type: a causal conclusion does not match a conditional conclusion just because both have a flaw. Next compare conditional direction and quantifiers. Then examine whether the choice uses the same premise-to-conclusion move.

If two answers remain, compare their validity. Is the original deductively valid, probabilistic, or flawed? Does it contain one conclusion supported by another intermediate conclusion? Does it rely on two premises that must be combined? A choice may match one visible phrase while missing the overall structure.

For a very dense stimulus, write a quick template. “All A → B; C is B; therefore C is A” is more useful than a detailed paraphrase. For a causal argument, use “after X, Y; therefore X caused Y.” For a sample argument, use “a subset showed X; therefore all members show X.” This makes comparison faster when the choices are unfamiliar.

Common parallel question traps

• [object Object]
• [object Object]
• [object Object]
• [object Object]
• [object Object]
• [object Object]

Mini practice: match the reasoning

Stimulus: “All contractors approved for the tunnel project completed a safety orientation. Lee completed the orientation. Thus, Lee was approved for the tunnel project.” The pattern is approved → orientation; orientation; therefore approved. The argument confuses a necessary condition with a sufficient one.

Choice A: “Every employee who completed payroll training can access the system. Noor can access the system, so Noor completed payroll training.” This is a parallel reversal: training → access; access; therefore training. Choice B: “Every employee who completed training can access the system. Noor completed training, so Noor can access it.” This uses the rule correctly and is valid. Choice C: “Some contractors completed orientation. Lee is a contractor, so Lee completed orientation.” This is a some-to-individual error. Choice D: “Lee completed orientation after the tunnel project was approved, so approval caused Lee's orientation.” This is a causal claim. Choice A matches the original flaw.

The best answer changes both the people and the activity, but preserves the inference pattern. The repeated words “orientation” and “contractor” are not what make an answer parallel. If choice A instead said every person with access completed training, the direction would be reversed in the premise and the argument would no longer share the same form.

A practice and review routine

Begin untimed. After reading the stimulus, write a one-line structure summary before looking at answer choices. Then label each choice's argument form. Once you can match accurately, add timing. If you repeatedly miss reversals, practice with arrows. If you miss quantifier shifts, mark the group and quantifier explicitly. If you choose answers by topic, force yourself to state the conclusion form first.

When reviewing an error, compare the stimulus and your chosen answer side by side. Name the exact mismatch: premise direction, quantifier, conclusion, or flaw type. Then explain why the credited answer matches those features. Do not only write “I should have read more carefully.” A specific correction provides a behavior to repeat.

Parallel reasoning can take longer than a short inference question because each answer contains a new mini-argument. Keep moving if no structure is emerging. Mark the question, answer the rest, and return if time allows. All questions count equally and wrong answers are not penalized, so do not leave an item blank.

Sources

• [object Object]
• [object Object]