Reading Comprehension curriculum

Find the sentence that says it, then check the option against that sentence word by word. The usual miss is an option that matches the sentence's topic but changes one detail, such as a number, a date, or who did what.

A correct paraphrase keeps the claim and changes the words; a trap keeps the words and changes the claim. Test an option by asking whether it would still be true if you knew only what the sentence means, not how it was phrased.

Before judging an option, mark the edges of the claim: who it covers, when, and under what conditions. A sentence about adult smokers in one city says nothing about teenagers or the country, even when an option makes that leap sound natural.

Note the quantifier every time. "Most" allows exceptions, "some" guarantees only a case or two, and "none" rules out every case. An option that turns "many" into "all", or "some" into "most", has changed the claim even if every other word matches.

Find exactly what the "not" modifies. "Not all of the samples were contaminated" means at least one was clean; it does not say most were contaminated, or that none were. Restate the sentence in positive form before comparing it with the options.

"If it rains, the game moves indoors" says nothing about what happens on a dry day. An option claiming the game stays outdoors when it does not rain reads a promise into silence. Keep the condition and the result together, and check which one the option asserts.

Before answering, replace each "it", "they" or "this" with the noun it stands for. Two nouns in the previous sentence may both fit grammatically, and the passage's logic, not the nearest noun, decides which one is meant.

"The study" or "these findings" points back to something specific, often several sentences earlier. When a passage mentions two studies, check which one the phrase refers to before crediting it with a result; options often attach a finding to the wrong study.

When a passage lists items across several sentences, count them as you go and note what each one is. Options often drop a member, add one from a neighboring list, or give the whole group a property that belongs to one item.

Identify both sides of every comparison and the dimension being compared: cheaper than what, faster by what measure. "A costs less than B to install" says nothing about which costs less to run, a substitution options make often.

Rebuild the order from verb tense and markers such as "before", "after" and "had already". The order of telling is not always the order of events: a passage can open with the outcome and explain its cause afterward.

Separate the passage's causal claims from its reports of timing or association. "Sales rose after the campaign" reports a sequence; "the campaign raised sales" asserts a cause. An option that upgrades the first into the second claims more than the text does.

For every claim, ask who holds it: the author, a cited researcher, or critics the author is about to answer. Many wrong options state a view the passage reports only so that it can reject it.

Read a claim at the strength the author chose. "The data suggest" is weaker than "the data show", and "may reduce" is weaker than "reduces". An option that drops the hedge, or adds one the author did not use, misstates the passage.

Words like "except", "unless" and "apart from" carve a group out of a general statement. Note what was carved out before answering, because the trap option usually applies the general rule to exactly the excluded case.

When the passage defines a term, its definition governs every later use of that term in the passage, even where your own sense of the word differs. Answer with the passage's meaning, and expect one option built on the everyday sense.

A passage often glosses a term in passing, with a dash, a parenthesis or "that is". Use that gloss rather than outside knowledge: the question tests whether you read it, and the passage may use the term more narrowly than a dictionary does.

Material in parentheses or between dashes can be removed without breaking the sentence, which makes it easy to skip. Questions sometimes ask about exactly that aside, such as a qualification, a date or an exception, so read it as part of the claim.

An example illustrates a point; it is not the point. If a passage says many cities cut traffic deaths and then describes one city's speed cameras, the claim is about the cities, and an option naming the cameras as the main idea mistakes the illustration for the thesis.

Read a number exactly as written, with its unit and qualifier. "Nearly 40 percent" is not "about half", and "rose to 12 percent" is not "rose by 12 percent". Rounding in your head is how a correct option starts to look wrong.

Per-person, per-year and per-unit qualifiers change what a figure means. The city with the most crimes may have a low crime rate per resident. Check that an option keeps the passage's unit before you compare its numbers.

The main idea is the claim the rest of the passage supports, so test a candidate by asking whether each paragraph serves it. An option that only one paragraph supports is too narrow; one that would fit many different passages is too broad.

The topic is what a passage is about, such as city trees. The thesis is what it says about that topic, such as that trees cool streets enough to justify their cost. Main-point questions want the thesis, and options that name only the topic are incomplete.

A good title covers the whole passage at the right scale. Reject titles that describe only the first paragraph, that promise more than the passage delivers, or that name the topic without the passage's angle on it.

A faithful summary keeps every main claim at the weight the passage gave it. Watch for summaries that add a claim the passage never made, drop a key qualification, or promote a minor point to the center.

An inference goes one step beyond the text, and the passage must supply the ground for that step. Point to the sentences that make the conclusion reasonable; if you cannot, the option rests on something you brought with you.

A must-be-true question allows no probabilities. The correct option follows from the stated facts alone, often by combining two of them. Any option that could be false in some situation the passage allows is out.

Here you rank options instead of proving one. Several may be possible, so choose the one that needs the fewest added assumptions. A modest option that follows from one sentence usually beats a bold one that needs three guesses.

An unwarranted leap needs a premise the passage never states. Find the missing link: if an option says a policy will save money, look for where the passage connects the policy to cost, not just to some other benefit.

An option can be true in the world and still be wrong, because the question asks what the passage supports. If you can defend an answer only with something you know from elsewhere, it is the trap, however accurate it is.

Match the strength of a conclusion to the evidence. A passage reporting that "early results suggest a benefit" does not support "the treatment works" or "researchers have shown". Strong verbs in an option are a signal to recheck the source sentence.

A finding about one group, place or period does not extend to all of them. If the passage studied college students in one country, an option about people in general has widened the scope beyond what the evidence covers.

"If a plant is a cactus, it stores water" does not mean every plant that stores water is a cactus. Reversing a conditional is one of the most common reading errors, so check which side is the condition before accepting an option.

"If you register early, you get a discount" does not mean late registrants get none; there may be other routes to a discount. Negating both sides of a conditional produces a claim the passage never made.

"Required for" marks a necessary condition: without it the result cannot happen, but having it does not guarantee the result. "Enough for" marks a sufficient one. Passages signal the difference with words like "only if", "must", "guarantees" and "all it takes".

An argument's unstated premise is the claim that bridges its evidence and its conclusion. To find it, set the evidence and the conclusion side by side and ask what would have to be true for one to lead to the other.

Separate what the author reports from what the author thinks. Evaluative words, concessions and the space given to each view reveal the author's position, even in a passage that mostly summarizes other people.

When a passage presents several views, look for the signals of endorsement: "rightly", "persuasively", or the author building on a view. A view introduced with "some argue" and followed by "however" is usually the one being resisted.

The next paragraph usually picks up the thread the last sentence opened: a question raised, a problem named, or a promise to examine something. Choose the option that follows that thread, not one that returns to a settled point.

To answer "what would follow if…", apply the passage's own reasoning to the changed condition. Use only the mechanism the passage describes; an option that needs a new mechanism goes beyond the text.

To apply a passage's principle to a new case, state the principle in one sentence first, then look for the case that shares its structure, not its subject. A case on the same topic with a different logic is the usual trap.

When a passage reaches a specific judgment, state the general rule it relies on. The right principle must justify that judgment without also justifying conclusions the passage rejects, so test it against both.

Given a stated principle, find the case that meets every condition it sets. Wrong options typically meet most of the conditions but miss one, often a qualifier tucked into the principle's last clause.

A passage's silence supports almost nothing. If an article on a vaccine's side effects never mentions cost, you cannot infer that the vaccine is cheap or expensive. Treat an omission as unknown unless the passage says its list is complete.

In a narrative, infer motive from what characters do and say, and from how the narrator frames them. Pick the motive that explains several details at once; an option that explains only the last action is usually too narrow.

Some passages leave the setting unstated and drop clues instead: the weather, tools, institutions, or how people speak. Reconstruct the situation from several clues together, since any single clue may fit more than one setting.

A passage can set up a comparison without stating it, for example by describing one school's results in detail and another's in a single sentence. Ask what the arrangement implies, but keep the conclusion within the details actually given.

When an answer follows from several steps, check each link in the passage. If A leads to B and B to C, the passage supports A leading to C; it does not support A leading to D just because D often follows C in the real world.

Test a candidate inference at the boundaries the passage sets: the smallest case, the largest case, and the stated exceptions. An inference that fails at one of those edges is not supported, even if it works in the typical case.

When the evidence fits more than one account, an option claiming the passage proves one of them overreaches. The supportable answer acknowledges the alternative, or names the evidence that would tell the accounts apart.

"Best supported" does not mean "the only possibility". Do not reject the right answer because you can imagine an exception; reject an option only when the passage supports a different one more strongly.

Some options sound like something the author would say, which makes them feel right. Ask which sentence makes this true. If the answer is the passage's general mood rather than a specific line, the option is unsupported.

Check every clause of an option separately. A common wrong answer is anchored in the passage for its first half and invented for its second, the two halves joined by "because" or "and therefore".

A restatement repeats something the passage says; an inference adds something the passage makes reasonable. When the question asks what can be inferred, an option that merely restates a sentence may be true, but it is often not what was asked.

When an answer feels obvious, name the assumption you supplied to reach it. If that assumption appears nowhere in the passage, the obviousness came from you, and the question is probably testing exactly that gap.

Find the claim the other sentences exist to support. Signal words like "therefore" and "so" help, but a conclusion can come first or sit in the middle. Test a candidate by asking "why should I believe this?" and checking that the rest of the passage answers.

Premises are the reasons offered for the conclusion. Background, examples and concessions can sit beside them without doing that work. A premise answers "why?" about the conclusion; a background sentence does not.

An intermediate conclusion is supported by one claim and then used to support another. In "Prices rose, so demand fell, so the store cut staff", the middle claim does both jobs. Questions about a sentence's role often hinge on spotting that double duty.

Describe what the sentence does, not what it says: is it the main conclusion, a premise, a concession, an example, or a view the author rejects? Then match that role, ignoring options that only paraphrase the sentence.

Name how the argument moves: by analogy, by counterexample, by eliminating alternatives, by applying a general rule, or from cause to effect. Describe it abstractly before reading the options, since they are usually written in abstract terms too.

A strengthener closes a gap between the evidence and the conclusion, often by ruling out another explanation. Find the gap first. An option that only repeats the conclusion, or supports a premise nobody doubted, does little.

A weakener attacks the link between evidence and conclusion, usually by showing another cause, a biased sample, or a case where the conclusion fails. It does not need to disprove the conclusion, only to make it less likely given the evidence.

A necessary assumption is a claim the argument cannot do without, and it is often modest, such as that a trend will continue or that two groups are comparable. Confirm it with the negation test: if denying it breaks the argument, it was necessary.

A sufficient assumption is a premise that, added to the argument, guarantees the conclusion. It is usually stronger than strictly needed. Look for the option that links the stated evidence directly to every term in the conclusion.

Turn a candidate assumption into its denial and reread the argument. If the argument falls apart, the assumption was necessary; if it survives, the assumption was merely helpful. Negate carefully: the denial of "all" is "not all", not "none".

Describe the flaw in your own words before reading the options: what does the argument take for granted, confuse, or ignore? Then match that description, instead of picking the first named fallacy that sounds familiar.

An ad hominem answers a claim by attacking the person who made it, their motives, character or situation. In a passage it may sound like "the critics work for the industry". That can raise a question about bias, but it does not show the claim is false.

A straw man restates an opponent's view in a weaker form and refutes that version instead. Compare how the passage describes the opposing view with what the opponent is quoted as saying; a gap between the two is the flaw.

A false dilemma presents two options as the only ones, so rejecting one seems to force the other. Ask what third possibility the passage ignored, such as a combination, a delay, or doing nothing at all.

A circular argument assumes what it sets out to prove, often in other words: "The policy is fair because it treats people justly." Check whether any premise is just the conclusion restated.

Composition assumes a whole has a property because its parts do; division assumes the parts share a property of the whole. A team of excellent players may not be an excellent team, and a large company can have small departments.

Equivocation uses one word in two senses, so the conclusion follows only by sliding between them. Track each repetition of a key term, such as "theory" or "natural", and ask whether it means the same thing every time.

Two things rising together does not show that one causes the other. Before accepting a causal conclusion, look for a third factor that could drive both, or for the possibility that the causation runs the other way.

Post hoc reasoning concludes that because B followed A, A caused B. A recovery after a new policy may come from something else that changed at the same time. Look for the comparison the passage did not make.

A conclusion about a group is only as good as the sample behind it. Ask how the sample was chosen and whether those who answered differ from those who did not, as when online survey respondents stand in for all voters.

A percentage cannot support a claim about numbers without its base. A product's market share can fall while its sales rise, if the whole market grew. Check whether the argument quietly switches between shares and counts.

Expert opinion is good evidence within the expert's field. An argument that cites a famous physicist on nutrition, or an unnamed "scientists say", borrows authority that does not cover the claim.

How many people believe something is evidence about belief, not about truth. "Millions use this remedy" shows the remedy is popular; it takes a separate argument to show the remedy works.

An argument from analogy holds only if the two cases are alike in the respect that matters. Find the property the conclusion depends on and check whether both cases share it; a difference there breaks the analogy, however similar the cases look otherwise.

"The new design is 30 percent more efficient" invites the question "than what?" A claim of improvement without a baseline cannot be evaluated, and an argument resting on one is incomplete.

Some arguments rest on a premise that, if true, undercuts their own conclusion. A passage that argues experts are unreliable by citing an expert's study has evidence and conclusion pulling against each other.

To find an argument with the same structure, reduce the original to a skeleton, such as "All A are B; C is A; so C is B", and compare skeletons, not topics. Options on the same subject are often the trap.

To match a flawed argument, name its flaw precisely first, for example "treats a necessary condition as sufficient". Then find the option that makes the same mistake, ignoring options that contain a different one.

When two speakers disagree, find the specific claim one asserts and the other denies. An option that only one speaker addresses, or that both would accept, is not the point at issue.

To find what two opponents share, look for a premise both rely on. They may disagree about the remedy while agreeing about the problem, and the correct option names that common ground.

A paradox question gives two facts that seem incompatible. The answer is a new fact that lets both be true, often by revealing a difference between the groups or periods involved. It explains both facts rather than denying either one.

When results differ from what you would expect, look for a factor that differs between the cases. If a hospital known for good care has a higher death rate, the explanation may be that it treats sicker patients.

The most useful question for judging an argument is one whose answer could strengthen or weaken it. Find the argument's key assumption and pick the question that tests it; facts about the general topic will not help.

Sort each sentence into what is claimed and what is offered as support. An argument can be full of true statements and still fail, if the statements offered as evidence do not bear on the claim.

Whoever makes a claim owes the evidence for it. Watch for passages that treat a lack of disproof as proof, such as "no one has shown this additive is harmful, so it is safe".

Tone comes from word choice, not from the topic: a passage about a disaster can be neutral, and one about cooking can be scornful. Collect the evaluative words and judge their direction and intensity before matching a tone word.

Tone can shift, with a passage opening admiring and ending skeptical. When asked about overall tone, weigh where the author ends up and how much space each attitude gets, not your first impression.

"Ambitious, if uneven" is praise with a fence around it. The qualifier often carries the author's real verdict, so an option calling the attitude simply enthusiastic, or simply critical, misses half the sentence.

Understatement downplays on purpose: calling a severe flood "a bit of a nuisance" signals how bad it was. Read it against the facts the passage gives; when the description is far milder than the events, the gap is the point.

Hyperbole exaggerates for emphasis, so "the line went on forever" means long, not endless. Read an overstatement for the attitude it expresses, and do not treat the exaggerated claim as a fact the passage asserts.

Irony says one thing and means the opposite. Signals include praise that contradicts the facts given, a mismatch between tone and content, and words like "naturally" or "of course" attached to something absurd.

Before choosing a tone, decide whether the author means the words literally. If an approving sentence follows details that make approval absurd, it is ironic, and options that take it at face value are wrong.

Sarcasm is pointed irony aimed at a target. Look for exaggerated praise, mock politeness, and quotation marks around a claim the author plainly rejects. Who the target is, and which way the jab points, matters more than the label.

Quotation marks around a word the author is not quoting signal distance: calling a change a "reform" in quotes tells you the author doubts it reforms anything. Read the marks as a comment on the word, not as emphasis.

Words like "seems", "arguably" and "tends to" lower the author's commitment. A hedged passage supports a cautious conclusion, and an option that states its claims as settled facts overstates it.

Words like "clearly", "undeniably" and "of course" raise the stated commitment, but they are not evidence. An author who boosts a claim is confident, and the claim still needs support from the passage.

Word choice can carry a judgment: "scheme" versus "plan", "bureaucrats" versus "officials". When an option calls the author neutral, check the nouns and verbs the author picked, because loaded words reveal the attitude.

A euphemism softens a harsh fact, as "workforce adjustment" does for layoffs. Recognizing one tells you what actually happened, and often how the writer, or a party the writer quotes, wants it to be seen.

A sudden move from formal to casual wording, or back, is usually deliberate. It can mark sarcasm, echo someone else's voice, or show the author dropping a pose. Ask what changed and why.

"Supporters claim the plan will cut costs" reports a view without adopting it. Verbs like "claim", "assert" and "insist" keep distance, while "show" and "demonstrate" grant credit. Options often treat a reported view as the author's own.

A concession grants a point in order to move past it: "Admittedly, the plan is expensive; still, it pays for itself." The author's position follows the "still" or "but", so read the concession as context, not as the view.

A rhetorical question asserts rather than asks. "Who would trust a forecast that has failed for ten years?" means no one should. Turn it into a statement to see the claim it makes.

Some comparisons exist to belittle, such as likening a policy to "rearranging deck chairs". The analogy expresses contempt and makes no literal claim, so the tone question is answered by what it ridicules.

Writers persuade through credibility (ethos), emotion (pathos) and reasoning (logos), and each leaves traces: credentials and fairness, vivid stories, or data and argument. Identify which a passage leans on by the kind of material it uses.

To find the purpose, look at what the passage does from start to finish: inform, argue, warn, entertain, or reconcile views. The verb in the right option matches the passage's overall action, not the action of one paragraph.

Vocabulary, assumed knowledge and examples reveal the audience. A passage that defines basic terms is written for newcomers; one that uses jargon freely assumes specialists. Pick the audience the passage's assumptions fit.

When asked what a word means "as used", set its most common sense aside and try each option in the sentence. The right choice keeps the sentence's meaning; a common sense of the word that does not fit is the usual trap.

Many words have several senses, and questions favor words whose usual sense is wrong in context, like "check" meaning restrain or "qualified" meaning limited. Read the surrounding sentence before you trust your recognition of the word.

Near-synonyms can share a meaning and differ in attitude: "frugal" approves where "stingy" criticizes. When options differ only in shading, choose the one whose attitude matches the author's.

A metaphor or simile works through one point of comparison. To read "the city's arteries clogged at rush hour", identify what is shared, here the flow of traffic, and do not carry over properties the comparison does not use.

Personification gives human agency to abstractions or objects, as in "the market punished the company". Read it for the point it makes, and notice when it hides the real actors, here the investors who sold their shares.

An allusion borrows meaning from a shared reference, such as calling a costly win a Pyrrhic victory. If you do not know the reference, the surrounding sentences usually signal its gist, so read them before guessing.

Deliberate repetition, such as the same phrase opening successive sentences, marks what the author wants remembered. When a phrase keeps recurring, it usually names the passage's central concern.

In a narrative, an early detail can prepare a later turn, such as a loose stair mentioned in passing before a fall. Questions about a detail's purpose often want this forward-looking role rather than its local meaning.

Tone options often include one at the wrong end of the scale, such as "enthusiastic" for a skeptical passage. Decide the direction first (positive, negative or neutral), then the intensity, and eliminate options on the wrong side before comparing the rest.

Ask what a paragraph does for the whole: introduce the problem, give evidence, answer an objection, or draw the conclusion. Questions about a paragraph's role want that function, not a summary of its content.

A topic sentence states a paragraph's claim, and it is usually, but not always, the first sentence. Find the sentence the others support; a paragraph can open with a question or an example and state its claim later.

Common patterns include chronological, comparison, cause and effect, and problem and solution. Name the pattern by the relationship between the parts, and check that it fits the whole passage, not just its opening.

Many passages state a claim first and then narrow to evidence and examples. Knowing that, you can expect the main idea near the top and treat the later specifics as support, not as new claims.

Some passages open with a case, a scene or an anecdote and only then draw the principle. Do not mistake the opening story for the point; the general claim that follows is what the story was for.

The controlling claim can open a passage, close it, or follow a setup paragraph. Locate it by asking which sentence the rest of the passage supports, wherever that sentence sits.

Authors often raise an opposing view before answering it, with phrases like "some argue" or "critics contend". Note where the response begins, and answer questions about the author's view from the response, not from the objection.

A common pattern is to state a view, concede what is right in it, refute the rest, and restate the author's position. Mapping these moves shows which sentences express the author's view and which only set it up.

Transitions announce how the next idea relates to the last: "however" turns, "moreover" adds, "therefore" concludes. Reading them closely lets you predict a paragraph's role before you read it in full.

"But", "yet", "whereas" and "although" signal a turn, and what follows a turn usually carries more of the author's weight. When a sentence pivots, read its second half as the claim.

"Furthermore" and "in addition" signal more of the same kind of support. They tell you the new material builds the same case rather than changing direction, so it serves the claim already made.

"Thus", "consequently" and "hence" mark a result or an inference. The sentence they introduce is often a conclusion, and the sentences just before it are its reasons.

"For instance" and "such as" introduce examples, and an example supports a claim made nearby. Questions often ask which claim an example serves, so look at the sentence just before it.

"First", "second" and "finally" lay out a roadmap. Count the promised items and check that you found each one, because a question may ask about the third reason when the passage makes it easy to lose track.

Writers link sentences by repeating key words, substituting pronouns or synonyms, and leaving out words the reader can supply. Following those links tells you which earlier idea a sentence continues.

Cohesion is surface linking, such as connectives and repeated words; coherence is whether the ideas actually fit together. A passage can put "therefore" between unrelated sentences, so judge the logic, not the signal word.

A digression leaves the main thread and returns to it. Recognizing one keeps you from treating its material as part of the central argument, and questions sometimes ask why the author included it at all.

An opening image or anecdote that returns at the end frames the passage. When the frame comes back, it is usually changed by what came between, and that change carries the author's point.

A closing paragraph can summarize, point to implications, call for action, or reframe the issue. Identify which it does, since an option describing a summary is wrong for a conclusion that calls for change.

Before the questions, jot one short clause per paragraph naming its job. The map takes seconds and answers most structure, main-idea and purpose questions without a full reread.

With two passages, read each for its claim, then compare: do they agree, disagree, or address different aspects? Note each author's method and tone as well, since questions ask how the second relates to the first.

Paired authors often agree on more than they first seem to. Find the specific claim one makes and the other rejects, and avoid options that overstate how far apart the two are.

Two authors who reach different conclusions may rest on the same assumption, such as that a problem is real. A question about what both would accept wants that shared ground.

Many stories move through setup, complication, climax and resolution. Placing an event in that arc tells you its role: an early detail may establish character, while a late one may resolve tension.

First-person narrators know only their own experience, third-person limited narrators follow one character's perceptions, and omniscient narrators see everything. In free indirect speech a character's thoughts appear in the narrator's voice, so a judgment may belong to the character, not the author.

In an experiment, investigators assign the exposure, such as a vaccine or a placebo, and compare the outcomes. In an observational study, they only record who was exposed and what happened. Experiments support causal claims more directly; observational findings leave more room for other explanations.

A control group shows what happens without the treatment, so the difference between groups can be credited to the treatment. Without one, any change over time could come from something else that changed at the same time.

People often improve when they believe they are being treated. Giving the comparison group a placebo, an inactive look-alike, separates the effect of the treatment itself from the effect of being treated at all.

When participants do not know which group they are in, their expectations cannot shape the results. Keeping the people who measure outcomes unaware as well, double-blinding, stops their expectations from shaping what they record.

Random assignment decides who gets the treatment by chance, so the groups start out alike, which supports cause-and-effect conclusions. Random sampling decides who is studied, so the sample resembles the population, which supports generalizing. A study can have one without the other.

A confound is a third factor linked to both the supposed cause and the effect. If coffee drinkers have more heart disease, smoking could be the confound if coffee drinkers also smoke more. Ask what else differed between the groups.

When two things go together, the causation may run the other way. People who exercise may be healthier because they exercise, or they may exercise because they are already healthy. Check whether the design rules this out.

A finding describes the people who were studied. If they were all volunteers from one university, the result may not hold for other ages, places or backgrounds, and an option generalizing to everyone overreaches.

When people choose whether to take part, those who choose may differ from those who do not, being more motivated, healthier, or more interested. Comparing volunteers with everyone else can mistake those differences for an effect.

Participants who drop out can change a study's result. If people who saw no benefit left early, the remaining group will look more successful than the treatment was. Check how many dropped out, and why.

Data that include only the cases that survived to be counted can mislead. Studying only companies still in business hides the strategies of those that failed. Ask which cases never made it into the data.

A result is established for a particular population, setting and time. A drug tested in adults says little about children, and a classroom finding may not hold online. Keep any conclusion inside those bounds.

A pattern observed within one range cannot be assumed to continue beyond it. If a fertilizer raised yields at low doses, the data say nothing about much higher doses, where the effect could level off or reverse.

A count needs its base. "200 cyclists were injured" means something different in a town of 5,000 riders than in a city of 500,000. Before comparing counts, ask how many people or events each came from.

A relative change compares two risks; an absolute change gives the actual difference. 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.

A percentage change depends on its starting point. A rise from 2 to 4 is a 100 percent increase, while a rise from 50 to 52 is 4 percent, though both added 2. Check what the percentage is measured from, and over what period.

Extreme results tend to be followed by less extreme ones, because part of any extreme result is luck. A school with unusually low scores one year will often improve the next with or without a new program, so a before-and-after comparison can credit the program for chance.

Studies often measure something easier than what they claim, such as test scores for learning or blood pressure for heart health. Check what was actually measured, and whether it is close enough to support the conclusion drawn.

Data people report about themselves are shaped by memory and by the wish to look good. People tend to underreport habits seen as bad and overreport those seen as good, so a survey of diet or exercise may not match actual behavior.

Statistical significance means a result would be unlikely if there were no real effect; it does not mean the effect is large. With a big enough sample, a tiny difference can be significant, so look for the size of the effect, not just whether it was detected.

A confidence interval gives a range of values consistent with the data. A wide interval means the estimate is imprecise, and an interval that includes no difference at all means the data cannot rule out that there is no effect.

One study can be wrong for many reasons, from chance to an error in method. A finding repeated by independent researchers with different samples is much stronger. Treat a single surprising result as a lead, not a settled fact.

Studies with striking positive results are more likely to be published than studies that find nothing. A set of published studies can therefore make an effect look stronger or more consistent than it is.

A vivid story makes a claim memorable, not more likely. One person who recovered after a remedy tells you nothing about how often the remedy works. Look for evidence on how many cases, compared with what.

An average describes a group, not any one member. If a program raises average test scores, some students may have gained a lot and others nothing. An option that applies the group result to one particular person overreaches.

One year's rise does not make a trend. Before accepting that something is increasing, check how many periods are involved, and whether the change is larger than the normal year-to-year variation.

When a passage describes a graph, read only what it states: values, direction, timing. "Rose steadily" and "rose sharply in the last year" describe different shapes, and an option should not add a peak or a dip the description does not give.

A count can rise while a rate falls, if the population grows faster than the events. More traffic deaths in a growing state can still mean a lower death rate per mile driven. Check which one the claim is about.

Who paid for a study is worth noting, because funding can shape which questions are asked and how results are presented. It does not by itself show the results are wrong; it is a reason to look harder at the method.

News summaries often state findings more strongly than the study did, dropping its hedges, its limits, or the population studied. When a passage reports both, rely on what the study itself found.

A too-broad option points in the passage's direction but covers more ground than it supports, such as "technology improves schools" for a passage about one tutoring program. Compare the option's scope with the passage's evidence.

A too-narrow option is true of one paragraph or detail and offered as the whole point. Test a main-idea option against every paragraph; if only one supports it, it is too narrow.

Some options are accurate statements about the passage that do not answer the question asked. Reread the stem, because an option can be true and still answer a different question.

An option that is true in the world but absent from the passage is wrong on a reading question. If your reason for choosing it is something you already knew, the passage has not supported it.

An opposite option reverses the passage's claim, often by one word: "increase" for "decrease", or a "not" added or dropped. Because it borrows the passage's vocabulary it can feel familiar, so check the direction of each claim.

A half-right option has one part the passage supports and one part it does not, often joined by "and" or "because". Judge each part separately; one unsupported half makes the whole option wrong.

Words like "always", "never", "must" and "proves" make strong claims. They are right only when the passage is that strong, so when the passage hedges, an extreme option overstates it.

An option that repeats exact phrases from the passage catches the eye even when it misstates the claim. Match the meaning, not the words, and be wary of an option that copies a sentence but changes its logic.

A scope swap applies the passage's claim to a different group, period or condition, such as findings about adults applied to children, or one year's results presented as a lasting pattern. Check every boundary the option sets.

A causal swap reverses cause and effect: the passage says stress led to poor sleep, and the option says poor sleep led to stress. Identify which way the passage's "because" runs.

A comparison swap reverses which side of a comparison is greater, or changes what is being compared. Note both sides and the direction when you first read the comparison, so a reversed option stands out.

An attribution swap gives the author a view the passage only reports, from critics or another researcher. Before matching an option, confirm whose view it states.

This option is an accurate, specific point from the passage presented as its central claim. It feels safe because it is true; reject it because the passage is not about that detail.

When asked for an argument's conclusion, beware the option that restates one of its premises. Premises are what the passage offers in support; the conclusion is what that support is for.

An option can compare two things the passage discusses separately but never compares. Unless the passage relates them, the comparison is unsupported, however reasonable it sounds.

Some options state what the author would likely say next. A question about what the passage says or implies wants the present text, not a prediction about the following paragraph.

An option can correctly answer a question about the passage, just not the one asked. Keep the stem's exact task in view, whether it asks for the purpose, the method or an assumption, while you check each option.

Read the stem before the options and take it literally: "the author would most likely agree" differs from "the passage states", and "primarily" asks for the main purpose. Restate the task in a few words before looking at the choices.

In an EXCEPT or LEAST question, four options fit and one does not. Mark each option yes or no against the condition and choose the odd one out; when reading quickly, it is easy to pick the first option that fits.

When a question lists several statements and asks which are supported, judge each statement on its own before looking at the combinations. Eliminating one statement often removes several answer choices at once.

Eliminate an option only for a reason you can name, such as the wrong scope, the wrong direction, or an unsupported claim. Rejecting on a feeling makes the right answer as likely to go as a wrong one.

When a question cites a line, read the sentences before and after it. The cited line often answers a question raised just before it, or is qualified just after, and that context decides its meaning.

Before reading the options, answer the question in your own words from the passage, then look for the option that matches. Prephrasing keeps attractive wrong options from setting your expectations.

Memory of a passage drifts toward what seemed plausible. Before choosing, return to the exact sentence that supports your answer and reread it, especially numbers, qualifiers, and who said what.

For your final choice, point to the words in the passage that make it true. If you can defend it only with an impression or an outside fact, look again; the best answer is the one you can prove from the text.