ACT Reading common mistakes
ACT Reading · Understand the traps and how to avoid them.
Choosing an answer because it repeats words from the passage rather than because it matches what the passage actually says.
Under time pressure the eye gravitates to familiar vocabulary, and the test writers deliberately seed distractors with verbatim phrases from the text while the correct answer is often a paraphrase.
What to do instead
Before looking at the choices, restate the answer in your own words, then pick the option that matches your paraphrase; eliminate any option whose only support is shared wording with the passage.
Selecting a true statement from the passage that does not answer the specific question asked.
Candidates confirm the option is factually supported and stop there, skipping the step of checking it against the exact wording of the stem (e.g., 'the main purpose' vs. 'a detail mentioned').
What to do instead
Underline the key word in the stem (main idea, primarily, most nearly, EXCEPT) and reject any choice that is accurate but off-target for that specific ask.
Answering detail questions from memory of a first read instead of returning to the lines.
Readers build a gist-level mental model and trust it, but the test hinges on precise wording that the gist drops (quantities, qualifiers, who did what).
What to do instead
Treat every line-referenced or specific-fact question as a lookup task: physically return to the cited lines or scan for the keyword, and verify the answer against the printed text.
Treating a stated opinion or hypothesis as an established fact when summarizing the passage.
Narrative and argumentative passages blend the author's claims, characters' beliefs, and reported findings, and hurried readers flatten all of these into 'the passage says.'
What to do instead
Track attribution while reading: note who is speaking or believing (author, a named researcher, a character) and make sure summary answers preserve that attribution.
Guessing at the meaning of a vocabulary-in-context word from its most common definition.
Candidates recall the dominant dictionary sense and ignore that the question asks how the word functions in this sentence, where a secondary sense is intended.
What to do instead
Substitute each candidate meaning back into the sentence and read the whole sentence; the correct sense must fit the surrounding logic, not just be a real definition.
Ignoring structural and punctuation signals that reveal an author's emphasis or shift.
Readers process for content and skim past transitions, colons, dashes, and paragraph breaks that mark contrast, concession, or the author's point.
What to do instead
During the read, circle contrast and cause-effect transitions (however, yet, therefore, in fact) and note what each paragraph is doing, not just what it says.
Misreading the author's tone or purpose as neutral when the language is evaluative.
Candidates focus on information transfer and miss connotative word choices, irony, or rhetorical framing that signal attitude.
What to do instead
Flag adjectives and verbs with a charge (e.g., 'claims,' 'admits,' 'remarkable') and ask whether the author is endorsing, doubting, or merely reporting before choosing a tone or purpose answer.
Evaluating an analogy or comparison by surface topic rather than by the relationship it expresses.
The two things compared usually come from different domains, and readers match on subject matter instead of on the underlying function the comparison is making.
What to do instead
State the relationship in abstract terms (X is to Y as cause is to effect, or part is to whole) and test each option against that relationship, not against the topic.
Judging a paired-passage question from one passage without integrating both viewpoints.
Candidates answer from whichever passage they read last or understood better, and treat the two passages as separate rather than as a comparison the question is testing.
What to do instead
For each paired question, write a one-line position for Passage A and Passage B first, then choose the option that correctly characterizes the relationship (agree, qualify, contradict) between them.
Accepting an inference that goes beyond what the passage supports, or rejecting one that requires a small logical step.
Candidates conflate inference with speculation in one direction and demand explicit proof in the other, so they either overreach or refuse a valid implication.
What to do instead
Apply the test 'must be true given the text': the correct inference needs direct textual support plus one defensible step, while any option requiring outside knowledge or a leap is wrong.