Elimination, Guessing, and Productive Review

An answer can identify the right topic and still make the wrong claim about it. Read a contender as a set of commitments: who or what, did what, under which conditions, with what degree of certainty? If a passage says that some seedlings grew faster under one laboratory treatment, an answer claiming that the treatment improves all plants outdoors changes the population and setting. Familiar nouns do not repair those changes. A single unsupported commitment can make the whole answer unsuitable, even when the rest accurately summarizes the text.

Use a short defect label to direct your check: wrong group, reversed relationship, added cause, changed certainty, or wrong task. These labels are questions to investigate, not labels to attach because an answer looks suspicious. 'All' is appropriate if the text supports all members of the stated group. 'May' can still introduce an unsupported possibility. An answer does not become defensible merely by sounding cautious. Compare the exact subjects and conditions before judging the strength of its verb.

Separate what the passage establishes from what could also be true. An experiment showing that a material tolerates heat does not by itself establish that the material is inexpensive, easy to manufacture, or suitable for every engine. You need not prove that an unsupported choice is impossible in the world. You need to show that it goes beyond the evidence required by this question. Conversely, do not reject a sound paraphrase just because it replaces the passage's vocabulary with different words.

Make elimination brief enough to help your pace. Once you can name a decisive defect, cross out that choice and direct attention to the remaining contenders. Bluebook's option eliminator can be undone if you discover that your reason was mistaken. [6] A crossed-out choice is a record of a judgment, not a fact. If the final two both fail, reconsider the earliest elimination instead of forcing one to fit. The worked example demonstrates why evidence for survival under one condition cannot carry a claim about a much larger natural process.

When two choices remain, stop asking which one sounds more convincing. State their difference in a form you can test. Does one explain a cause while the other merely report a sequence? Does one compare two objects while the other describe only one? Does one supply the main verb while the other leave only a participle? A useful comparison identifies the smallest consequential difference, then returns to the relevant evidence. Repeatedly reading both choices without naming that difference rarely creates new information.

For reading questions, compare each contender with the precise claim, relationship, or role requested. For editing, insert each into the complete sentence and test the disputed structure. For synthesis, translate the goal into a short checklist before comparing. If the student wants to compare materials and purposes, a choice that accurately compares materials alone is incomplete. Truth and task fulfillment are separate checks. The longest choice can fail by adding an unsupported detail; the shortest can fail by omitting a required part.

Keep the comparison small without making your reading too narrow. To decide between two inferences, you may need a qualification in the first sentence as well as the result in the last. To choose a transition, you may need the claim that a new example develops, not merely the immediately preceding noun. Read the part that can settle the difference, and widen that reading when a pronoun, exception, or causal link depends on earlier context. A prediction made before reading the choices remains provisional.

Sometimes you cannot settle the final two within a reasonable first-pass effort. Choose the answer with better support, mark the question if a specific later check seems useful, and move on. Having two uneliminated choices does not establish that they are equally likely to be correct; you may have stronger evidence for one, or may have eliminated the correct answer. On return, perform the missing check rather than restart the entire struggle. In the worked example, the decisive difference is whether both requested comparison dimensions appear.

An answer change should follow a changed understanding of the evidence or rule. Useful reasons include noticing a qualifier, correcting a reversed comparison, finding the sentence's actual subject, or recognizing that a chosen answer fulfills only half of the goal. 'I have looked at this too long' and 'this answer suddenly feels strange' identify discomfort, not a defect. Ask yourself: What did I miss, and how does it make the current answer worse than the replacement? You do not need a written explanation during the test; one precise mental sentence can be enough.

Give the replacement its own test. Finding a flaw in A does not automatically prove B. If you discover that A describes only one of two studies, check whether B actually compares the studies accurately rather than merely mentioning both. In editing, reconstruct the sentence with the new choice, including punctuation already outside the blank. A correction that fixes agreement but creates a fragment is not complete. The object is to improve the answer, not simply to move away from the first selection.

Keeping an answer can also be an active, evidence-based decision. Suppose your original choice says that a finding 'supports' a hypothesis and the text reports evidence consistent with it. A rival claiming that the finding 'proves' the hypothesis may overstate the conclusion. If a targeted reread confirms that distinction, keep the original choice and finish the review. Do not demand perfect certainty about every question before moving on. Once the relevant check supports the selection, repeated reopening can consume time needed elsewhere.

Review your changes after practice in both directions: wrong to right, right to wrong, and wrong to wrong. Record the reason you used at the moment, then ask whether that reason was valid. A fortunate switch based on letter patterns is still an unreliable process; an evidence-based change deserves inspection even when the replacement remains wrong. Look for the missing second check. The worked example shows a justified switch prompted by one previously overlooked word, although, which changes the status of an entire clause.

A review flag identifies unfinished work; it does not tell you how valuable a return will be. At the end of a module, first verify that an answer is selected for every question. College Board states that there is no penalty for guessing. [5] A guess can still be wrong, but leaving a question unanswered gives up the opportunity to answer it correctly. The flag itself is not a selected response. You can revisit questions in the current module, but you cannot return after leaving that module. [2]

Once every question has a selection, prioritize returns with a specific unresolved check and a reasonable time cost. 'Check whether I used the correct table row' is more actionable than 'this passage felt hard.' A suspected copied value or missed negation may be worth checking before a low-confidence vocabulary choice for which you have no new approach. Confidence matters as a signal, but the strongest review target is often the one with an identifiable, repairable defect. Do not invent point values for different questions to justify the order.

Use the clock to limit the scope of the return. If a check should take a brief reread, perform that check; if it expands into a complete restart, reconsider whether the remaining time supports it. Reserve time to confirm the selected answer after making a change. Do not remove a provisional answer while deciding what to replace it with. If the module is almost over and you cannot resolve a question, keep or select the best-supported remaining choice. With no usable evidence, select an available choice rather than leave the item blank.

Avoid two end-of-module traps. First, do not change letters to make an answer pattern look balanced; your selections on other questions do not establish the answer to this one. Second, do not interpret two crossed-out choices as proof of a 50-50 situation. Your eliminations and remaining evidence may be uneven or mistaken. Use any surviving evidence, then finish. The review order below is a training recommendation under stated constraints, not an official scoring formula or a guarantee that the most difficult question deserves the most time.

Keep these checks in mind

• A half-right answer can fail through one changed group, condition, relationship, or degree of certainty.

• For the final two, name the deciding difference and test it against the passage, sentence, data, or synthesis goal.

• A justified answer change requires a defect in the current choice and support for the replacement; keeping an answer also deserves evidence.

• First ensure that every item has a selected answer, then prioritize short, specific checks of suspected errors.

• Crossed-out choices, subjective confidence, and earlier answer-letter patterns do not determine correctness.

Begin with four short drills, one per lesson, then complete twelve multiple-choice questions, three per lesson. Labels distinguish SAT-format questions from strategy decisions. Give the evidence or stated condition that decides each item. Model drill responses are examples; another well-supported response may also work.