Parallel Reasoning questions ask you to find the answer choice with the same logical structure as the stimulus — and the most common mistake is searching for the answer choice about the same TOPIC instead.
If the stimulus argues from a general rule to a specific conclusion about penguins, the correct parallel answer might be about tax policy, cooking, or astronomy — the topic is irrelevant. What has to match is the underlying logical skeleton: how many premises, what type of claims they make, and how the conclusion follows from them.
Your brain naturally gravitates toward answer choices that feel thematically similar to the stimulus, because topical similarity is an easier, faster pattern to notice than structural similarity. Test-writers know this, and often place a topically-similar-but-structurally-wrong answer choice specifically to catch that instinct.
Strip the stimulus down to its logical skeleton before looking at answers: how many claims are there, are any of them conditional (if/then), does the conclusion follow necessarily or only probably, is there a causal claim, a comparison, a generalization from a sample. Do the same reduction for each answer choice, and match skeleton to skeleton.
| Stimulus feature | What to check in each answer choice |
|---|---|
| Number of premises | Same count of supporting claims? |
| Conditional logic present? | Does the answer also use if/then structure? |
| Strength of conclusion | Necessarily true, or just probable/likely? |
| Causal vs. correlational claim | Same type of claim being made? |
Parallel Reasoning is one of the more learnable LR question types precisely because it isn't really testing your reasoning about the topic at all — it's testing whether you can abstract an argument's shape away from its content, a skill that transfers identically across every parallel question you'll ever see.