MCAT CARS common mistakes
MCAT CARS · Understand the traps and how to avoid them.
Summarising each paragraph in your own words as you read, then answering questions from that paraphrase rather than returning to the passage text.
Paraphrasing feels like active reading and saves time, but the MCAT CARS answer choices are engineered to punish paraphrase drift: a subtle shift in scope, degree, or attribution between the passage and your summary becomes invisible once the original wording is gone.
What to do instead
Read for the author's argument structure (claim, evidence, concession, conclusion) and mark where each element sits, but when a question hinges on wording, re-read the exact sentence and its neighbours before choosing. Treat your summary as a map, never as the evidence.
Treating the passage as a neutral information report and missing that the author is arguing a position, so you answer 'what does the passage say' instead of 'what does the author think'.
Many CARS passages are adapted from essays and academic prose where the thesis is stated indirectly through hedges, concessions, and rhetorical questions; candidates trained in science reading look for data, not stance, and default to a neutral voice.
What to do instead
On the first read, force an explicit answer to: what is the author trying to get me to believe, and what is she arguing against? Note the contrast words (however, yet, although, admittedly) - they usually mark the author's real position rather than the foil she is describing.
Confusing an example, illustration, or cited study inside the passage with the author's main point, then selecting an answer that restates that detail.
Details are concrete and memorable while the thesis is abstract, so memory favours the example. Test writers also place a detail-restating option that is factually true of the passage, which feels safe.
What to do instead
Before answering any main-idea or function question, ask what work the detail does - support, concession, counterexample, or transition. If an option is true but only covers one paragraph's evidence, it is a supporting detail, not the main point.
Choosing an answer because it is a reasonable real-world belief or matches your prior knowledge of the topic, rather than because the passage supports it.
CARS passages often touch on familiar debates (education, technology, ethics), and familiar framing triggers prior beliefs. Options written as sensible generalities are attractive precisely because they need no textual support.
What to do instead
Apply a strict provenance test: point to the sentence or sentences that license the option. If your justification is 'that's generally true' or 'that sounds right', eliminate it. Outside knowledge is never required and frequently is the trap.
Ignoring qualifier and degree words (some, most, all, may, must, primarily, only, tends to) when comparing answer choices, treating 'often causes' and 'invariably causes' as interchangeable.
Skilled readers skim function words for speed, and the stem rarely flags that degree is the discriminating feature. Two options can share the same idea and differ only in strength, so the choice feels arbitrary.
What to do instead
When two options survive, compare them word by word for scope and strength, then check which level of commitment the passage actually makes. Extreme language is not automatically wrong, but it requires explicit textual warrant.
Answering 'the author would most likely agree/disagree' questions by imagining what a sensible person with that view would think, instead of extrapolating from the author's stated commitments.
These items invite projection: the stem asks about a hypothetical scenario the passage never mentions, so candidates reason from general plausibility rather than from the author's established principles, tone, and stated values.
What to do instead
First state the author's core commitments in one sentence, then test each option against them: does it follow from those commitments, contradict them, or is it simply unaddressed? Unaddressed is not the same as agreed.
Treating 'which of the following, if true, would most weaken/strengthen the argument' as a comprehension question and picking an option that merely restates the passage's claim or a background fact.
These items require modelling the argument as a claim plus an assumption, a skill distinct from reading comprehension; without an explicit assumption, every option looks like it could be relevant, and restatements feel supportive.
What to do instead
Write the argument's conclusion and its unstated assumption before reading the options. The correct strengthener supplies the assumption; the correct weakener attacks it. Reject options that are merely consistent with the conclusion without bearing on the reasoning.
Spending a uniform amount of time on every passage and every question, so a single dense humanities passage consumes the clock and forces rushed guesses on later, easier passages.
CARS passages vary widely in difficulty, but candidates often treat finishing each passage as a commitment, sunk-cost style, rather than managing the section as a whole. Anxiety about leaving questions blank compounds the effect.
What to do instead
Set a per-passage time budget and practise abandoning an over-invested question after a fixed interval with a flagged best guess. Drill full 90-minute sections so pacing decisions become automatic rather than deliberated under pressure.
Reading the questions first and then hunting the passage for matching keywords, producing answers built from isolated sentences stripped of their context.
This strategy is borrowed from fact-retrieval tests and feels efficient, but CARS questions test argument-level relationships; keyword matching reliably finds the sentence that mentions a term while missing the concession or qualification surrounding it.
What to do instead
Read the passage fully first, then answer questions with the passage's structure in mind. When you do locate a line, read the sentence before and after it, since authors routinely qualify a claim immediately after stating it.
Leaving the hardest questions blank or letting the timer expire with unanswered items, rather than making a reasoned elimination-based guess.
Candidates fear that guessing lowers accuracy or violates some implicit rule, or they lose track of time and never return to flagged items. On a scored multiple-choice section, a blank is a guaranteed loss with no compensating benefit.
What to do instead
There is no penalty for wrong answers, so every question gets an answer. Practise elimination to a 50/50 and commit; keep a strict flag-and-return routine and reserve the final minutes of the section for confirming that no item is unanswered.