How to Use Elimination on a Quantitative Reasoning Question

The short answer: eliminate contradictions, then verify

Use elimination as a screen: translate the question into conditions every correct answer must satisfy, remove only choices that break one, then calculate or substitute if more than one remains. Name what is requested and its unit. Check the operation, sign, and any range the wording requires. An estimate can narrow options, but cannot settle a close choice. The aim is to avoid calculations that cannot change the decision while keeping the answer checkable. A hard constraint follows from the problem itself. If the prompt asks for a whole-number count, a fraction fails; if it asks for a distance, an answer in units of time fails. Such a choice is impossible under the stated conditions. A heuristic is a useful guide, such as expecting a result near a rounded estimate. It can make an option less plausible, but does not prove the option wrong. A surviving choice has only passed the checks you made. It has not automatically been verified. For example, imagine a question asks for the total after adding the same amount to several equal groups. Before looking at choices, record the number of groups, the amount in each, and whether anything is later added or removed. A candidate that reverses addition and removal may fail the prompt; one with the right scale may still use the wrong operation. Eliminate the first only when the wording makes the order clear. Treat scale as a clue, then work the relationship or substitute the candidate to decide. This is a compact version of planning, monitoring, and evaluating a method. In “Thinking Aloud to support mathematical problem-solving,” the Education Endowment Foundation describes making strategy choices visible, checking progress rather than blindly continuing, and evaluating the method after solving. That page discusses scaffolding pupils’ mathematical problem solving; it does not test this exact answer-elimination sequence or show that it improves adult test performance. The practical extension is modest: make each rejection explainable, and stop eliminating when the evidence is only a hunch.

Sources: Thinking Aloud to support mathematical problem-solving

What elimination can prove—and what it cannot

Elimination proves that a particular answer choice conflicts with a condition in the question. It does not prove that a surviving choice is correct unless the stated conditions leave only that choice or a calculation confirms it. This distinction matters because a choice can feel unlikely without being impossible. A sound rejection names the rule the choice breaks; a guess based on appearance, familiarity, or inconvenience does not. Quantitative reasoning means using quantities and their relationships to answer a question. A hard constraint is a condition every valid answer must satisfy, such as having the requested unit or falling within a range that follows from the prompt. An estimate is an approximate value or interval used to judge scale. A constraint rules out contradictions; an estimate narrows possibilities. If two options fit, the estimate has not decided between them. Consider an illustration: a question asks for the number of whole boxes needed to hold 43 books if each box holds 10 books, with every book packed and no box overfilled. The answer must be a whole number, and 4 boxes cannot hold all 43 books because their capacity is 40. Those are direct contradictions. Five boxes meet the capacity requirement, but the conclusion depends on the question asking for the number of boxes required, rather than asking how many boxes remain empty or how many books are in each box. The assumptions make the bound valid: books are counted as whole objects, each box has capacity 10, and all books must fit. Elimination establishes that fewer than five is impossible; the requested quantity still needs checking. If a problem says a distance is positive, a negative answer contradicts the condition. If an estimate suggests a distance is around 20 units, an answer of 19 may be plausible and 200 may be implausible, but neither conclusion is exact unless the wording or calculation supports it. State the failed condition and verify that it follows from the question. The Education Endowment Foundation’s “Thinking Aloud to support mathematical problem-solving” describes planning, monitoring, and evaluating as ways to make a chosen approach explicit and check progress. This offers a model for checking whether an operation matches the question and whether its result makes sense. It does not show that this particular elimination sequence improves test performance. The What Works Clearinghouse guide “Improving Mathematical Problem Solving in Grades 4 Through 8” recommends monitoring and reflection for grades 4–8; it does not validate this tactic for adults or show effects on IQ. A raw item outcome is what happened on an item under its conditions, such as selecting the correct answer; alone, it is not a normed score or a statement about ability. For this question, use elimination to identify contradictions, then solve or substitute when choices remain. A quick rejection is evidence about one option’s fit to the prompt, not proof about the answer or the test taker.

Sources: Thinking Aloud to support mathematical problem-solving; Improving Mathematical Problem Solving in Grades 4 Through 8

Why translate the prompt before scanning the choices?

Name the requested quantity before looking at the choices, then restate the relationships that produce it. This keeps the options from silently choosing the operation for you. A compact equation or sentence gives you something to check: does this answer represent the quantity asked for, in the right unit, after the stated steps? Translation makes the conditions for a valid answer visible, so elimination tests those conditions instead of rewarding whichever option looks familiar.

The What Works Clearinghouse guide, Improving Mathematical Problem Solving in Grades 4 Through 8, recommends helping students monitor problem solving and use visual representations. This guidance concerns school instruction, not a trial of answer-choice elimination or adult test performance. Applied cautiously, the principle is to make the setup inspectable: represent what the words say, then check whether an answer fits. It does not show that a shortcut improves scores.

Consider an illustration. A pantry has seven trays with twelve sealed jars on each tray. A worker removes nine jars. How many remain? Before scanning hypothetical choices, write the sequence: seven groups of twelve, then subtract nine. In symbols, 7 × 12 − 9. The requested quantity is jars remaining, not jars per tray or the number removed. Multiplication represents equal groups; subtraction follows because removal happens after the jars are counted. Subtracting nine from twelve first would describe a different operation.

Build the checks in order. First, represent the relationships: 7 × 12 − 9. Second, set a bound. Seven trays contain 84 jars, and removing nine leaves fewer than 84 but more than zero. This assumes each tray starts with twelve jars and only nine are removed. Third, estimate: 84 minus about 10 is about 74, so an option near 75 is plausible; a negative answer or one above 84 conflicts with the setup. The estimate narrows possibilities but does not establish an exact answer. Finally, compute: 84 − 9 = 75 jars. Check each hypothetical option against the equation and unit. Values outside the bound fail; among values inside it, only 75 satisfies the calculation.

This example shows why choices should come after translation. A number such as 63 might tempt someone to multiply seven by nine, while 84 reflects stopping before the removal. Both arise from nearby quantities in the wording, but neither answers the full question. The equation preserves the sequence and exposes what a distractor may have omitted. Matching a number to one phrase is weaker than matching it to the whole relationship.

OpenStax's Prealgebra section, 5.6 Ratios and Rate, explains that a ratio preserves which quantity is first and which is second, and that measurements in different units must be converted before forming a ratio. This matters when a prompt asks for a comparison: ‘blue tiles to red tiles’ is not interchangeable with ‘red tiles to blue tiles.’ Reversing the order changes the ratio. OpenStax supports these mathematical checks, not claims about test speed or elimination effects.

Read for the requested result, not merely for numbers. Note the target and unit, express the relationship in words or symbols, then inspect options. If a condition is unclear, do not invent one to reject a choice. Elimination is justified only when a choice contradicts what the prompt states.

Sources: Improving Mathematical Problem Solving in Grades 4 Through 8; 5.6 Ratios and Rate

Which constraints can safely eliminate an option?

A useful order is to check the answer’s unit and direction, then its sign, whether it must be a whole number, any divisibility condition, a justified bound, and finally its scale. Eliminate only when the prompt entails the condition. Units act as a basic consistency check. If the question asks for distance, an answer in hours fails unless it asks for a rate or derived measure. In a rate such as kilometres per hour, numerator and denominator describe different quantities, and order matters. OpenStax’s “5.6 Ratios and Rate” distinguishes ratios, which compare quantities with the same unit, from rates, which compare different units, and explains that unlike measurement units must be converted before forming a ratio. This supports checking setup, not rejecting an option just because its unit looks awkward. Translate what is compared and what is requested. Direction is often just as cheap to check. If the prompt asks how many items remain after some are removed, an answer larger than the starting amount conflicts with that operation, assuming nothing is added. If change is defined as final minus initial, a decrease has a negative change. A decrease does not mean the final value is negative: direction of change and sign of the result differ. Match the option to the operation and requested quantity. Next ask what kind of quantity the answer represents. An explicitly requested count of indivisible objects is a whole-number quantity, so a choice such as 6.4 objects can be rejected. A measured length, average, probability, or rate can be fractional. An average of 6.4 books per shelf does not claim that any shelf holds a fraction of a book. Apply the integer check only to a whole-object count. Divisibility applies only if equal whole-number groups with no remainder are required; unequal groups, leftovers, or an average change the condition. Bounds can eliminate options too, provided their assumptions are visible. Suppose a question states that a positive whole-number quantity is divided among four equal groups and asks for the amount in each group. A total of 18 implies 4.5 per group, which is incompatible with the stated whole-number condition; if equal whole-number groups are required, the total must also be divisible by four. If the question instead asks for an average across four groups, 4.5 may be perfectly valid. The arithmetic is unchanged; the stated condition changes what is valid. Bounds also follow from operations: subtracting a positive amount from a starting total cannot produce a larger result, without additions or wraparound rules. State those conditions. Finally, scale is a screening check, not a proof. If a calculation multiplies two positive quantities greater than one, the product exceeds either factor; an option far below both deserves scrutiny. But scale depends on sign, units, and operation: multiplying by a fraction below one reduces a positive quantity, while subtraction can produce a negative result. Identify these facts first. The What Works Clearinghouse’s “Improving Mathematical Problem Solving in Grades 4 Through 8” recommends monitoring problem solving and teaching visual representations in grades 4–8. It does not study adult multiple-choice elimination. Apply the broader advice modestly: make a condition visible, test each option against it, and retain choices where the condition is uncertain. Eliminate only an unavoidable contradiction.

Sources: 5.6 Ratios and Rate; Improving Mathematical Problem Solving in Grades 4 Through 8

When does estimation help, and when does it mislead?

Estimation helps when it separates possibilities by more than the uncertainty introduced by rounding. It is a screen: it can rule out a value that is plainly too large or too small, or show that a computed answer is in the right neighborhood. It becomes misleading when a rounded result is treated as if it were exact. If two answer choices sit close together, the estimate may support both. If the estimate still fits multiple choices, calculate or substitute into the original question to verify. Imagine a question asks for 18% of 245, with choices 43, 44, 45, and 46. Rounding 245 to 250 and 18% to about one fifth gives an estimate near 50. That flags a result far from the low forties. But it does not distinguish 43 from 44, and the rounded inputs both move the result. A more controlled estimate keeps one source of approximation visible: 10% of 245 is 24.5, and 8% is 19.6, so the total is 44.1. That points toward 44, but if the task requires choosing between 44 and 45, the estimate alone still does not prove which is correct. Exact calculation gives 0.18 × 245 = 44.1; if choices are integers and the prompt asks for a rounded whole number, the rounding instruction determines the answer. Without that instruction, neither integer exactly equals the result. The wording must also say whether to round. A useful way to make uncertainty explicit is to bound the estimate rather than announce a single rounded number. If you replace 245 by a value known to be within five, and 18% by a value known to be within one percentage point, calculate a low and high result using those endpoints. The resulting interval describes what the approximations permit; it is not a formal confidence interval. If several choices fall inside it, the interval has not decided among them. Tighten the bounds, use exact arithmetic, or test each survivor in the original relationship. OpenStax’s “Solve Proportions and their Applications” models estimation as a reasonableness check alongside a solved proportion: its examples compare the exact result with a nearby easy calculation or a range. Estimates check scale; the equation supplies the exact result. Rounding can mislead in multiplication, division, and percentage change, where an input difference or wrong base can shift the result. OpenStax also defines a percent proportion by relating the amount to its base, so identify that base before estimating. “Predictions & Estimates” from The Math Learning Center makes a related instructional point: pausing to estimate can help a learner attend to context before choosing an operation, but prediction is not a contest to get closest to the exact answer. Its account is a practitioner reflection about classroom mathematics, not a controlled test of multiple-choice elimination, so it supports the purpose of a pause rather than a performance claim. Note what was rounded and which operation the prompt requires. Keep the estimate when it quickly detects an impossible scale. When options cluster, when approximation bounds overlap choices, or when wording affects the base or rounding rule, let exact calculation decide. An estimate narrows; it does not certify.

Sources: Predictions & Estimates; 6.5 Solve Proportions and their Applications

How do you handle rates, ratios, and percentages without a shortcut trap?

For a ratio, preserve the quantities in the order named; for a rate, track which unit belongs above and below the fraction; for a percent, identify the base before comparing choices. Reversing one of these relationships can produce a plausible number that answers a different question. OpenStax’s “5.6 Ratios and Rate” defines ratios as comparisons in the same unit and rates as comparisons across different units. “6.5 Solve Proportions and their Applications” stresses keeping the comparison order consistent. These are useful elimination checks once the wording has been translated correctly.

Consider an example: a recipe uses 2 cups of oats for every 3 cups of flour. How much flour goes with 8 cups of oats at the same ratio? The requested comparison is oats to flour, 2 to 3. Since 8 is four times 2, flour must be four times 3, or 12 cups. A choice of 5⅓ cups may come from reversing the ratio and calculating 8 × 2/3. That calculation fits flour-to-oats, but not the direction asked. Write labels beside the numbers to eliminate the reversal. Then substitute: 8:12 simplifies to 2:3, confirming the candidate preserves the stated ratio.

Rates need a unit check. Imagine a cyclist travels 42 kilometers in 3 hours and the question asks for kilometers per hour. Divide 42 kilometers by 3 hours to get 14 kilometers per hour. A choice of 0.071 hours per kilometer is the reciprocal rate, with different units, so it answers a different question. If choices show numbers only, attach the requested unit mentally or in a note. A value that seems plausible in size is not enough to show the direction is correct. OpenStax’s ratio-and-rate chapter keeps units in the rate and defines a unit rate as one with a denominator of one unit.

Percent questions add a base-selection trap. “What is 20% of 50?” asks for 20/100 of a base of 50, giving 10. “10 is what percent of 50?” asks for 10/50, giving 20%. For an increase from $50 to $60, the change is $10, measured against the original $50 base: a 20% increase. A tempting 16⅔% uses the new $60 as the base and answers another comparison. Eliminate it only if the prompt asks for increase from the original amount. OpenStax’s proportion chapter expresses amount divided by base as percent divided by 100, keeping the base explicit.

A proportional check can narrow choices quickly. If 4 notebooks cost $10 at a constant unit price, 8 should cost $20; a choice of $18 does not preserve the same dollars-per-notebook rate. But a discount, fixed fee, or tiered price can break that direct relationship, so do not assume proportionality without support in the prompt. OpenStax describes a proportion as equality between ratios; the mathematics supports checking an equal relationship, not assuming every situation has one.

Use a short sequence: mark the quantities and their order; attach units; identify the percent base; then test whether the candidate preserves the relationship. A wrong order, incompatible unit, or incorrect base can establish a contradiction. Familiar magnitude alone cannot. If multiple choices still fit, calculate the exact comparison or substitute each into the original wording. That keeps elimination as a screen and lets the stated relationship decide the answer.

Sources: 5.6 Ratios and Rate; 6.5 Solve Proportions and their Applications

What should you do when several choices survive?

When two or more choices survive the safe checks, stop crossing out answers and ask what is the smallest calculation that separates them. “This would take longer” is not a mathematical contradiction. Keep a choice until the prompt’s relationships show it cannot be right; then calculate, substitute, or compare the remaining candidates directly. Consider an example: a notebook costs $80, and its price is reduced by 15 percent. What is the sale price? Suppose the choices are $12, $68, $72, and $92. A quick range check rules out $92 because a reduction cannot make the price exceed the original $80. But that does not settle the answer. The $12 choice is tempting because it is the discount amount, not the sale price. The $68 and $72 choices both look possible as prices below $80. Guessing that “15 percent is about one tenth” might make $72 seem plausible, but that approximation is too coarse to distinguish the survivors. Now do only the calculation that resolves the ambiguity. Fifteen percent of $80 is 0.15 × 80 = $12. The question asks for the amount after the reduction, so subtract: $80 − $12 = $68. The wording controls the final operation. The result is in dollars, is below the original price, and differs from the discount itself. The arithmetic verifies $68; the initial range check did not. This is a useful place to compare candidates against the original relationship rather than repeat a whole solution. If a choice says $72, test it: the reduction from $80 would be $8, which is 10 percent, not 15 percent. If a choice says $12, it answers “how much was taken off?” rather than “what is the sale price?” The distinction is not that one number is absurd. It is that each candidate must satisfy both the calculation and the requested quantity. OpenStax’s *Solve Proportions and Their Applications* models this general check: solve the relationship, substitute into the original equation, and see whether the quantities and units still fit. The stopping rule is simple: if your exact constraints leave one answer and the prompt’s relationship confirms it, stop. If two choices remain, identify the operation or term on which they differ and compute that part. If substituting a candidate is quicker, put it into the original relationship. If neither test decides the issue, solve exactly. This follows the plan-monitor-evaluate approach described in the Education Endowment Foundation’s *Thinking Aloud to support mathematical problem-solving*: inspect whether the chosen method is working and change it when it is not. That source describes instructional practice in school mathematics; it does not test this multiple-choice tactic or establish that it saves time for adults. After choosing, make one final check against the question’s wording and units. Ask whether you found the amount removed, the amount remaining, a rate, or a total; then verify that the number carries the requested unit. This catches a correct calculation attached to the wrong question. For example, $12 is correct as the discount and wrong as the sale price. Elimination has done its job when it narrows the work; verification finishes it.

Sources: 6.5 Solve Proportions and their Applications; Thinking Aloud to support mathematical problem-solving

What can strategy practice change—and what does that say about a score?

A higher score after practice has several possible explanations: familiarity with the format, better use of a test-taking strategy, stronger knowledge in the tested subject, or improved performance on similar tasks. A score records performance under particular conditions; alone, it cannot identify the cause or establish that general intelligence changed. “Near transfer” means improvement on tasks similar to the trained task; “far transfer” means improvement on substantially different abilities or outcomes. *Retesting in Selection: A Meta-Analysis of Practice Effects for Tests of Cognitive Ability* synthesized 50 studies, 107 samples, and 134,436 participants in organizational and educational settings. Scores tended to rise when people retook cognitive-ability tests: the adjusted overall effect was 0.26. Effects were larger when practice included coaching and when the same forms were reused. The abstract calls for more research on how retesting affects validity interpretations. It does not isolate elimination practice or identify gains as increased general ability; the repository notes its copy may differ from the final journal version. A broader educational-test synthesis likewise found score gains, while leaving their cause open. *The Impact of Test Preparation on Performance of Large-Scale Educational Tests: A Meta-analysis of Experimental Studies* combined 28 experimental and quasi-experimental studies of students taking large-scale educational tests. Compared with no preparation or usual practice, preparation was associated with improved performance (g = .26, 95% CI .10 to .42). Workbooks, socio-affective strategies, and test-taking instruction moderated the effect; the review found little evidence for practice tests or sample items alone. The authors suggest domain knowledge or test-specific skills could explain gains. These school-based interventions do not test this elimination routine or establish an effect on adult IQ. Cognitive-training research asks whether practice on one task carries over to others. *Near and Far Transfer in Cognitive Training: A Second-Order Meta-Analysis* combined results from prior meta-analyses. Its working-memory models included 99 near-transfer and 119 far-transfer estimates; another model added 233 far-transfer estimates from other programs. The analysis found near transfer to memory tasks, with variation by population, while far-transfer effects were small or null, particularly after accounting for publication bias and placebo effects. It covered varied programs, not quantitative elimination, and notes uncertainty in smaller evidence groups and bias adjustments. Together, these findings support a limited conclusion: preparation can change measured performance, and some gains may remain close to the practiced format or skill. They do not show that learning an elimination check raises general intelligence. Better performance on fresh questions using the same operation would indicate learning that operation; alone, it would not establish far transfer. Use a practice result as feedback about performance under its conditions. Test IQ Free offers an adaptive 50-question practice test in pattern reasoning, number patterns, and spatial reasoning. Its estimated IQ range comes from an unvalidated conversion with no validated adult reference population; it is not a normed comparison or established statement about a person’s ability. The optional $9 report adds explanation, not validity. Review which operation needs work, practice it on fresh examples, then see whether the method helps with similar problems.

Sources: Retesting in Selection: A Meta-Analysis of Practice Effects for Tests of Cognitive Ability; The Impact of Test Preparation on Performance of Large-Scale Educational Tests: A Meta-analysis of Experimental Studies; Near and Far Transfer in Cognitive Training: A Second-Order Meta-Analysis

What should you do on your next question?

On a quantitative question, use this sequence. Write what the answer must name, including its unit. State one hard condition from the wording, such as a required whole-number count, and a range a plausible answer should occupy. Cross out a choice only when it contradicts a condition that follows from the prompt. If several choices survive, calculate or substitute into the original relationship; do not reject one because it looks awkward or takes longer to check. After choosing, note which condition decided the answer and whether your check was exact or approximate. The note shows whether the error was in translation, a condition, or an estimate. Improving Mathematical Problem Solving in Grades 4 Through 8, a What Works Clearinghouse guide, supports monitoring and reflecting during problem solving. It does not test this tactic or adult performance. A meta-analysis abstract in Cornell’s repository reports that practice effects on cognitive-ability tests varied with conditions such as coaching and repeated forms; it does not show that this method changes general intelligence. For low-stakes practice, try Test IQ Free’s adaptive 50-question test in pattern reasoning, number patterns, and spatial reasoning, then review the worked answers. Its estimated IQ range comes from an unvalidated conversion without a validated adult reference population. For consequential education, employment, disability, or clinical decisions, seek assessment from a qualified professional.

Sources: Improving Mathematical Problem Solving in Grades 4 Through 8; Retesting in Selection: A Meta-Analysis of Practice Effects for Tests of Cognitive Ability

Questions readers ask