Reading questions reward one discipline above all: answer from the passage, at the strength the passage states. Almost every wrong answer breaks that rule in a recognizable way. It says more than the text, less than the text, or something the text never said. The chapters move from reading what is stated to inferring one step beyond it, then to arguments, tone, structure and research claims, and end with how wrong answers are built and how to rule them out.
Each chapter opens with the short version. Tap one to read the detail.
Read exactly what is said
~2 min
Most errors start with misreading a sentence, not with deep reasoning. Track quantifiers, negations, conditions and who is speaking, read numbers with their units, and judge the main idea by what every paragraph supports.
Literal questions look easy, which is why they cost points. The common mistake is reading a sentence for its gist and then matching an option that shares the gist but changes a detail. Small words carry the claim. "Most" allows exceptions and "all" does not. "Not all of the samples were contaminated" means at least one was clean. "If it rains, the game moves indoors" says nothing about dry days. "The data suggest" is weaker than "the data show."
Attribution matters as much as wording. Passages report views in order to answer them, so before matching an option, ask whose claim it is: the author's, a researcher's, or the critics'. Numbers need the same care. "Rose to 12 percent" is not "rose by 12 percent," and the city with the most crimes may have a low rate per resident.
For the main idea, separate the topic from the thesis. City trees are a topic; the claim that trees cool streets enough to justify their cost is a thesis. Test a candidate against every paragraph: if only one paragraph supports it, it is too narrow, and if it would fit almost any passage on the subject, it is too broad.
Rule: before comparing options, restate the relevant sentence in your own words with every quantifier, negation and qualifier intact.
Infer one supported step, no further
~2 min
An inference goes one step past the text, and the passage must supply the ground for it. Most wrong inferences overreach in degree or scope, reverse a conditional, or lean on outside knowledge.
An inference is a conclusion the passage makes reasonable without stating it. The test is whether you can point to the sentences that support it. If your defense depends on something you know from elsewhere, the option is a trap, however true it is in the world.
Overreach is the usual failure. A passage reporting that early results suggest a benefit does not support "the treatment works." A study of college students in one country does not support a claim about people in general. Conditionals cause a second family of errors: "if a plant is a cactus, it stores water" does not mean every plant that stores water is a cactus. It also does not mean a non-cactus stores none.
Question wording sets the bar. "Must be true" means the option follows from the stated facts alone. "Most strongly supported" means ranking the options and choosing the one that needs the fewest added assumptions. "Best" does not mean "only," so do not reject an answer because you can imagine an exception.
When an answer feels obvious, name the assumption you supplied to get there. If it appears nowhere in the passage, the obviousness came from you, and that gap is often exactly what the question tests.
Rule: accept an inference only if you can name the sentence that grounds it and the inference claims no more than that sentence supports.
Map the argument before judging it
~2 min
Find the conclusion, find the premises offered for it, and locate the gap between them. Strengthening, weakening, assumption and flaw questions all ask about that gap.
Every argument question starts the same way. Find the conclusion by asking which claim the other sentences exist to support, then find the premises, the reasons offered for it. Background, examples and concessions often sit beside the premises without doing their work.
Between the premises and the conclusion there is almost always a gap, and most question types are about it. A strengthener closes the gap, often by ruling out another explanation. A weakener widens it, usually with another cause, a biased sample, or a case where the conclusion fails. An assumption is what the argument needs to cross the gap. Test a candidate by denying it: if the argument collapses, the assumption was necessary.
Flaws are recurring ways of crossing the gap badly. A flaw named only in the answer choices is easy to mismatch, so describe it in your own words first. Ad hominem attacks the arguer instead of the claim. Post hoc reasoning infers a cause from sequence alone, as when a team wins after changing its uniforms and the uniforms get the credit. Appeals to authority cite someone outside their field. Begging the question assumes the conclusion it set out to prove.
The trap in this chapter is choosing an option that is true and relevant to the topic but does not touch the gap.
Rule: state the conclusion and its main premise in one line each, then judge every option by what it does to the link between them.
Tone lives in word choice
~2 min
Tone comes from evaluative words, hedges, boosters and figures of speech, not from the subject. Decide the direction first, then the intensity, and check whether the author means the words literally.
A passage about a disaster can be neutral, and one about cooking can be scornful, because tone comes from word choice rather than topic. Collect the evaluative words ("scheme" versus "plan," "bureaucrats" versus "officials") and judge their direction before choosing a tone word.
Two judgments decide most tone questions. First, direction: positive, negative or neutral. Eliminate the options on the wrong side of the scale, since one is almost always placed there on purpose. Second, intensity: "ambitious, if uneven" is qualified praise, and the qualifier often carries the real verdict.
Then check whether the author means the words literally. Irony says the opposite of what it means; watch for praise that clashes with the facts given, or "of course" attached to something absurd. Quotation marks around a word the author is not quoting signal distance. Hedges like "arguably" lower commitment, and boosters like "clearly" raise it without adding evidence.
Reported views are the biggest trap. "Supporters claim the plan will cut costs" reports a view without adopting it, and a concession such as "admittedly, it is expensive" grants a point only to move past it. The author's position usually follows the "still" or "but."
Rule: decide the direction, then the intensity, then whether the author means it literally, and answer only after all three.
Know what each paragraph is for
~2 min
Structure questions ask what a paragraph or sentence does, not what it says. A one-clause map of each paragraph answers most of them, and transition words tell you where the argument turns.
Before the questions, jot one short clause per paragraph naming its job: introduces the problem, gives evidence, raises an objection, answers it, concludes. The map takes seconds and answers most structure, main-idea and purpose questions without a full reread.
Transitions announce how ideas relate. "However" and "yet" turn, and what follows a turn usually carries more of the author's weight. "Moreover" adds support of the same kind, "therefore" marks a conclusion, and "for instance" introduces an example of the claim just made. A common pattern is to state an opposing view, concede part of it, refute the rest, and restate the author's position. Knowing the pattern tells you which sentences are the author's view and which only set it up.
Do not trust connectives blindly. Cohesion, the surface linking, is not coherence. A passage can put "therefore" between sentences that do not follow from each other, so judge the logic, not the signal word.
With paired passages, read each for its claim, then compare. Authors who disagree about a remedy may agree about the problem, and options that overstate how far apart they are are a frequent trap.
Rule: describe what a sentence or paragraph does in one verb phrase before reading the options about it.
Read study claims at their real strength
~2 min
Passages that report research invite overreading. Check the design, the comparison group, who was studied, and what was actually measured, and read numbers with their base.
In an experiment, investigators assign the exposure, such as a vaccine or a placebo, and compare outcomes. In an observational study they only record who was exposed and what happened, which leaves more room for other explanations. A control group shows what happens without the treatment. A placebo separates the treatment's effect from the effect of being treated at all, and blinding keeps expectations from shaping the results.
The usual mistake is accepting a causal conclusion from a correlation. Ask what else differed between the groups: coffee drinkers with more heart disease may also smoke more. Ask whether the causation could run the other way. And ask who was studied, because a finding about volunteers at one university does not cover everyone.
Numbers mislead in predictable ways. A count needs its base: 200 injuries means different things among 5,000 riders and among 500,000. A risk that doubles from 1 in 10,000 to 2 in 10,000 has risen 100 percent in relative terms but only by 1 in 10,000 in absolute terms. Statistical significance means a result would be unlikely if there were no real effect; it says nothing about whether the effect is large.
Rule: before accepting a research claim, name the design, the comparison, the population and the base behind every number.
How wrong answers are built
~2 min
Wrong options follow a small number of patterns: too broad, too narrow, true but irrelevant, half-right, reversed, or built on outside knowledge. Answer in your own words first, then eliminate only for a reason you can name.
Wrong answers are built to attract readers who skim. A too-broad option points in the passage's direction but claims more ground, and a too-narrow one offers a single paragraph's detail as the whole point. A true-but-irrelevant option is accurate about the passage and answers a different question. A half-right option is supported in one clause and invented in the other, often joined by "because." Reversals swap cause and effect, flip a comparison, or give a critic's view to the author, while reusing the passage's own words so that they feel familiar.
Two habits defeat most of them. First, prephrase: answer the question in your own words before reading the options, so attractive wording cannot set your expectations. Second, eliminate only for a stated defect, such as wrong scope, wrong direction, or unsupported. Rejecting on a feeling throws out right answers as often as wrong ones.
Read the question stem literally. "The author would most likely agree" differs from "the passage states," and in an EXCEPT question four options fit and you want the one that does not. When a question cites a line, read the sentences around it.
Rule: for your final choice, point to the words in the passage that make it true; if you cannot, look again.
Written by Keentune. We are not affiliated with or endorsed by the organizations whose documentation informs this guide, and any linked sources belong to their respective owners.
All exam, test, and product names and trademarks are the property of their respective owners and are used here for identification and reference only. Keentune is independent study practice — not affiliated with, authorized, or endorsed by any of these organizations.