How to Solve Arithmetic Word Problems on a Reasoning Test

What is the reliable first move on an arithmetic word problem?

Before choosing an operation, identify what the question asks and how the amounts relate. Modeling means expressing quantities and their relationship; operation choice follows, then calculation. In “A jar held 12 counters. Then 5 more were added. How many are in it now?”, the unknown is the ending amount. Starting amount plus change equals ending amount, so 12 + 5 = 17. The question asks for the total now, not the change. If the same numbers asked for a missing starting amount, seeing “more” alone would not tell you the operation. The article “Theoretical-methodological approaches and textbook design: analysis of arithmetic word problems in Spanish textbooks” distinguishes understanding quantity relationships from selecting operations by textual cues. For your next item, name the unknown and state the relation; write an equation if it clarifies it. This routine is not a validated way to raise an IQ score. An answer shows performance on that problem under its conditions, not a person’s worth or fixed potential.

Sources: Theoretical-methodological approaches and textbook design: analysis of arithmetic word problems in Spanish textbooks

Why can a simple calculation become a difficult story problem?

A story problem asks you to do more than calculate. You must understand the sentence, work out what the quantities mean in relation to one another, choose a mathematical operation, and carry it out. A miss can happen at any of those points. Someone may know how to subtract but still select addition because a familiar word suggests it; someone else may identify the right subtraction and make an arithmetic slip. The answer is therefore not simply a measure of calculation skill.

A useful model separates four connected jobs: read the question, build a representation of the situation, plan the calculation, and execute it. A representation identifies which amounts are known, which is requested, and how they fit together. The jobs can overlap: an awkward calculation may prompt a re-reading, which changes the model. Still, separating them helps locate where an attempt went off course without turning one response into a judgment about ability.

An adult experiment, “The interaction of linguistic and arithmetic factors affects adult performance on arithmetic word problems,” tested how selected features changed response time. Twenty-five students solved 320 one-step problems in German. The researchers varied addition versus subtraction, whether a calculation required carrying or borrowing, a more noun-heavy wording called nominalization, and lexical consistency. Lexical consistency means that a cue word’s usual operation hint agrees with the operation required by the relation in that problem. Each of the four factors affected response times, and some language and arithmetic factors interacted when they were conceptually linked. The outcome was time on these tasks; the study did not test a teaching routine, an IQ assessment, or whether practice changes general intelligence.

The design helps explain why ‘just do the arithmetic’ misses part of the work. A carrying calculation can take more effort even when the story relation is unchanged. Noun-heavy wording can make the same structure less direct to parse. A cue can also pull attention toward an operation that does not answer the question as phrased. These are different sources of friction: calculation, language understanding, and the point where words meet the mathematical relation. The experiment’s small sample and German one-step tasks are a narrow basis; its response-time findings are not a score rule for every adult or test.

A separate study, “The Demands of Simple and Complex Arithmetic Word Problems on Language and Cognitive Resources,” examined 444 students in grades three through six from seven schools in the western Netherlands. Students completed 48 paper-and-pencil problems, including symbolic arithmetic, story problems, two-step problems, and problems with an irrelevant number. Researchers analyzed answer correctness. Within each grade, average performance did not differ across problem types; reading comprehension, however, was more important for two-step story problems than for one-step problems. The result suggests comprehension may matter especially when a problem has multiple steps. It does not show that language alone determines an individual answer, and participants were children, not adults taking an IQ test.

Together, these studies suggest a more useful review question than ‘Am I bad at arithmetic?’ Ask which job failed. Did the wording obscure the question? Were the quantities represented in the wrong relationship? Did the operation fail to match that relationship? Or was the model right and the arithmetic wrong? This diagnosis is a practical inference from the studies’ separation of task features and solution demands, not an intervention they tested. It can direct practice toward reading, modeling, operation choice, or calculation. The evidence describes why items can be harder; it does not establish that this routine improves adult IQ-test performance.

Sources: The interaction of linguistic and arithmetic factors affects adult performance on arithmetic word problems; The Demands of Simple and Complex Arithmetic Word Problems on Language and Cognitive Resources

Why do keywords fail when the unknown changes position?

A word such as ‘more’ describes a relationship; it does not, by itself, tell you which operation answers the question. The same relation can call for addition or subtraction depending on which quantity is known and which is missing. Consider two original examples. In the first, a blue ribbon is 5 centimeters longer than a red ribbon, and the red ribbon measures 12 centimeters. To find the blue ribbon, add: blue = red + 5 = 12 + 5 = 17 centimeters. In the reversed question, the blue ribbon is still 5 centimeters longer than the red one, but the blue ribbon measures 17 centimeters. To find the red ribbon, subtract: red = blue − 5 = 17 − 5 = 12 centimeters. The relationship has not changed; the unknown has. Write the relation before calculating: B = R + 5. If B is missing and R is known, evaluate 12 + 5. If R is missing and B is known, rearrange the same relation as R = B − 5. The equation makes the direction of the comparison explicit, so the arithmetic follows from the quantities rather than from the word. If you treat ‘longer’ as an instruction to add in both questions, the second answer becomes 22, which cannot fit the sentence: a red ribbon of 22 centimeters would be longer, not shorter, than a blue ribbon of 17. The 2024 article “Theoretical-methodological approaches and textbook design: analysis of arithmetic word problems in Spanish textbooks” describes additive word problems as change, compare, combine, or equalize structures, with different subtypes depending partly on the unknown quantity. A 2025 study, “How Keywords Impact Word-Problem Performance: When ‘More’ Does Not Mean to Add,” offers a bounded example of the risk. Researchers examined equations constructed for four word problems containing “more” by 112 third-grade students in a U.S. Southwest school district. Most students made accurate equations on the two consistent problems, where addition matched the cue, while fewer did so on the two inconsistent problems, where the required operation did not match the assumed cue. The study concerns schoolchildren and four specific items, so it cannot tell us how adults perform on reasoning tests. Adult task evidence points in the same practical direction without testing an instructional shortcut. In “The interaction of linguistic and arithmetic factors affects adult performance on arithmetic word problems,” 25 students solved 320 one-step problems in German; the experiment measured response times as linguistic and arithmetic features varied. Ask, ‘Which amount is known, which is missing, and what sentence would still be true if I inserted my answer?’ Then choose the operation that makes that sentence hold. This is a way to make one solution more checkable, not evidence that a particular routine raises adult IQ-test scores or general intelligence.

Sources: How Keywords Impact Word-Problem Performance: When ‘More’ Does Not Mean to Add; Theoretical-methodological approaches and textbook design: analysis of arithmetic word problems in Spanish textbooks; The interaction of linguistic and arithmetic factors affects adult performance on arithmetic word problems

What changes when the story asks for a comparison, a rate, or a missing part?

The relation changes, so the same operation cue may not work across problem types. A comparison relates two amounts by a difference; an equal-groups problem relates the number of groups to the amount in each; a part-whole problem asks how parts make a total or what part is missing. Describe that relation first, then write an equation. The important question is which quantity is unknown, not which arithmetic word appears.

Suppose one shelf holds 4 more books than another, and the larger shelf holds 13. How many are on the smaller shelf? Let L be the larger amount and S the smaller. The relation is L = S + 4. Since L is known and S is requested, S = 13 − 4 = 9 books. “More” describes the gap from smaller to larger; it does not mean add when finding the smaller amount. Check by restoring the relation: 9 + 4 = 13.

For equal groups, suppose 3 boxes each contain 6 pencils. The total is 3 × 6 = 18 pencils. If the total and number of boxes are known instead, and the question asks for pencils per box, calculate 18 ÷ 3 = 6 pencils per box. Both questions use the relation total = groups × amount per group, but the unknown changes. “Each” may help you notice the structure; the known and requested quantities tell you which calculation answers the question.

A missing-part example has another arrangement: there are 15 counters in all, 9 blue, and every counter is either blue or red. How many are red? Write total = blue + red, so red = 15 − 9 = 6 counters. Check by adding the parts: 9 + 6 = 15. The condition matters: if other colors were possible, the given information would not determine the red count.

These are not three unrelated tricks. They show why relating quantities is more dependable than assigning an operation to one word. “Theoretical-methodological approaches and textbook design: analysis of arithmetic word problems in Spanish textbooks” distinguishes additive structures, including change, compare and combine, from multiplicative structures such as rates and equal groups. It also explains that the unknown’s position can change the operation required. Its analysis concerns selected Spanish third-grade textbooks, so it helps organize problem structures; it is not evidence that adults improve IQ-test performance by using a particular routine.

For practice, keep labels attached to quantities: larger and smaller, number of groups and amount per group, or total and known part. State exactly what is missing before calculating. A story about distances or packages can share the same relation as one about books or pencils, even though its surface details differ. Recognizing the relation lets you apply a method across those stories and check whether the answer fits. It does not show broad reasoning ability or predict performance on every intelligence test; it gives you a clearer way to solve this kind of item and locate whether a mistake involved the relation, operation choice or calculation.

Sources: Theoretical-methodological approaches and textbook design: analysis of arithmetic word problems in Spanish textbooks

How do you handle two steps without losing track of the question?

Keep the requested quantity in view, then name any amount you must find on the way. Give each calculation a job: find the intermediate amount, label it, and use it to answer the original question. This helps prevent an intermediate result from being mistaken for the final answer and makes a mistake easier to locate. It is a way to organize a solution, not a demonstrated method for raising adult IQ-test scores. Imagine this original example: A club has 4 tables, with 6 chairs at each table. It borrows 5 more chairs. How many chairs does it have altogether? The requested quantity is the total after borrowing. First find the equal groups: 4 tables × 6 chairs per table = 24 chairs. Label 24 “chairs before borrowing.” Then add the borrowed chairs: 24 + 5 = 29 chairs altogether. The label matters: 24 is correct arithmetic, but it does not answer the question. Stopping there may mean the solver omitted the second relation, not that the multiplication was wrong. A compact chain is enough: chairs at tables = 4 × 6 = 24; final chairs = 24 + 5 = 29. Units help test whether the operations fit the story. Estimate as a quick check: four groups of six is just under 25, so adding five should put the total just under 30. The result, 29, is plausible. An inverse check gives another view: remove the five borrowed chairs from 29, then divide 24 among four tables to recover six per table. These checks test consistency, though they cannot prove that the story was modeled correctly. Two-step items may combine reading demands with arithmetic demands. “The Demands of Simple and Complex Arithmetic Word Problems on Language and Cognitive Resources” studied 444 students in grades three through six in the Netherlands using paper-and-pencil arithmetic tasks, including one-step and two-step word problems. Reading comprehension was more important for performance on two-step than one-step word problems. This is a finding about school-age children and the study’s tasks; it does not test the routine above or establish an effect for adults. Nor does it show that a word-problem result measures general intelligence. “The interaction of linguistic and arithmetic factors affects adult performance on arithmetic word problems” measured response times for 25 students solving 320 one-step problems in German. The researchers manipulated operation, carrying or borrowing, nominalized wording, and whether a cue word agreed with the needed operation. Each factor affected response time, with interactions between some language and arithmetic factors. Because these were one-step problems, the study does not directly test two-step strategies. Taken together, the findings support a limited point: solving a story problem can require coordinating its wording and arithmetic, while extra steps can add further demands. Neither study establishes a guaranteed technique for faster or more accurate adult test performance. Under time pressure, label only the intermediate result that keeps the chain clear, then continue to the requested quantity. During review, ask whether you misread the story, found the intermediate amount incorrectly, or stopped before answering the original question. That diagnosis points to a specific step to practice, without turning one missed item into a judgment about fixed ability.

Sources: The interaction of linguistic and arithmetic factors affects adult performance on arithmetic word problems; The Demands of Simple and Complex Arithmetic Word Problems on Language and Cognitive Resources

When should you write an equation or draw a diagram?

Use the smallest representation that makes the relation clearer. For a short change or comparison problem, a labeled equation may be enough. A bar, table, or sketch can help when parts or the unknown’s position are hard to hold in mind. Ask not whether a picture looks clearer, but what relation it would make visible. If it adds nothing beyond the sentence and equation, it costs time without helping. This separates a picture of the story from a model of its mathematics. A decorative drawing shows the objects mentioned; an organizational diagram shows how quantities fit together: which parts make a whole, which amount is larger, or how a change links a starting amount to an ending amount. In “There are 18 tickets in all; 11 are used; how many remain?”, a bar split into a used part and an unknown remainder can show the part-whole relation. The equation 11 + x = 18 states it more compactly. The sketch earns its place only if seeing the parts helps identify the unknown. “Theoretical-methodological approaches and textbook design: analysis of arithmetic word problems in Spanish textbooks” distinguishes organizational illustrations, which represent mathematical structure, from pictures that merely reproduce the situation. Its review reports prior student research in which representational pictures did not improve success, while structure-focused diagrams were associated with better problem solving. The article analyzes selected Spanish primary-school textbooks; it does not test adults taking timed reasoning tests. It supports the distinction between repeating story details and exposing a relation, not a claim that sketches raise adult test performance. The limits matter. In “Problem-appropriate diagram instruction for improving mathematical word problem solving,” an adapted multiple-baseline study gave staggered instruction in line diagrams, tables, and graphs to 67 Japanese junior-high students. Researchers reported increased use of appropriate diagrams and improved problem-solving performance across three problem types. Perceived cognitive-load changes varied by timing and problem type; the authors noted that practice could contribute to decreases in reported load. This was instruction in how and when to use representations, not a test of whether adults should draw during a reasoning test. Diagram use is a learned choice, not an automatic advantage. The What Works Clearinghouse’s 2021 “Assisting Students Struggling with Mathematics: Intervention in the Elementary Grades” rates well-chosen concrete and semi-concrete representations, mathematical language, and deliberate word-problem instruction as supported recommendations for elementary intervention. The guide concerns students through grade six; it does not validate a timed adult strategy or establish an IQ effect. It supports teaching a representation for a purpose rather than adding pictures indiscriminately. On a test item, try an equation first when the relation is clear and the unknown can be named directly. Draw only when the sketch will separate parts, mark a comparison, or keep an intermediate amount visible. Before drawing, ask what it will show that the sentence or equation does not. If there is no concise answer, skip it; if there is, draw only that structure and calculate.

Sources: Assisting Students Struggling with Mathematics: Intervention in the Elementary Grades; Theoretical-methodological approaches and textbook design: analysis of arithmetic word problems in Spanish textbooks; Problem-appropriate diagram instruction for improving mathematical word problem solving

What does instruction research support—and what does it leave open?

Instruction research supports teaching word-problem structure directly, using clear mathematical language, and choosing representations that help learners see relationships. The What Works Clearinghouse practice guide, Assisting Students Struggling with Mathematics: Intervention in the Elementary Grades, was released in 2021 and addresses elementary intervention, listing grades kindergarten through 6. It rates recommendations for systematic instruction, clear mathematical language, well-chosen concrete or semi-concrete representations, deliberate word-problem instruction, and timed fluency activities as having “strong evidence.” These ratings support instructional choices in the guide’s school context. They do not test the relation-first routine described here as a way to raise adults’ performance on an IQ test. Evidence for teaching children mathematical ideas cannot be converted into evidence about adult scores, and fluency practice is not evidence of increased general intelligence. The guide’s recommendations fit parts of a word problem that learners can make visible. Clear language can help a student state what quantities mean; deliberate instruction can make their relation explicit; and a suitable representation can keep parts, groups, or comparisons organized. In a relation-first routine, these are practical functions, not claims that one script has been validated. A learner can label known and missing quantities, then write an equation that preserves the story’s relation. A sketch may help when it clarifies that relation; if the equation already shows the unknown and how quantities connect, a picture may add nothing. The test is whether the representation helps explain what each quantity means and why the operation fits. The Springer article “An empirical analysis of additive and multiplicative structures in primary textbooks” describes categories such as change, comparison, equal groups, and rate in selected Spanish third-grade textbooks. It helps explain why instruction may address relationships rather than a list of cue words: problem structures connect quantities in different ways. But the article analyzes textbooks for young pupils, not adult test-taking methods. It offers language for describing structures, not proof that adult practice produces broader gains. For adult self-practice, the modest conclusion is useful. Treat the routine as a way to make a solution checkable: state the relation, choose a representation only when it clarifies that relation, calculate, then test whether the answer fits. Practice varied examples with feedback so review can distinguish a reading or modeling error from an operation choice or arithmetic slip. Improvement on practiced forms may reflect learning those operations or familiarity with their formats. That matters, but it is narrower than far transfer: improvement on substantially different tasks or a broad ability. These sources do not establish that this routine raises adult IQ-test scores, changes general intelligence, or predicts performance beyond practiced problems. Use it to learn and diagnose arithmetic reasoning, not as a promise of a higher score. The limits belong beside every practical inference.

Sources: Assisting Students Struggling with Mathematics: Intervention in the Elementary Grades; Theoretical-methodological approaches and textbook design: analysis of arithmetic word problems in Spanish textbooks

How should you review an error or work under a time limit?

Classify the miss before practicing: did you misread the question, misstate the relation, choose an unsuitable operation, or calculate incorrectly? On a timed item, keep the check compact: name the requested quantity, state the relation, write a short equation, estimate the answer’s size, and check its unit. This is a practical heuristic, not a tested speed intervention.

Consider an example: one amount is 7 more than another; the larger is 22, and the question asks for the smaller. Adding 7 because you saw “more” models the relation backward: smaller amount + 7 = 22. That is an operation-choice error. If the equation fits but 22 − 7 is worked incorrectly, the likely failure is computation. If the answer is 15 boxes when the question asks for items per box, revisit the requested quantity and units. These examples are illustrative, not evidence about error frequency. “The interaction of linguistic and arithmetic factors affects adult performance on arithmetic word problems” reports an experiment measuring response times for 25 students solving 320 one-step problems. Cue consistency and arithmetic features affected response times, but the study did not test this review routine or a timed-test strategy.

Use the smallest representation that makes the relation clear; stop adding notes once the unknown and relation are visible. Then practice the failed step with fresh problems and inspect worked answers. “Assisting Students Struggling with Mathematics: Intervention in the Elementary Grades” gives strong-evidence ratings to deliberate word-problem instruction, clear mathematical language, representations and timed activities in its elementary intervention context. This does not establish that timed practice improves adult reasoning-test scores. Test IQ Free offers low-stakes practice; its estimated IQ range comes from an unvalidated conversion with no validated adult reference population, so it is neither normed nor a confidence interval. The optional paid report adds explanation, not validity.

Sources: Assisting Students Struggling with Mathematics: Intervention in the Elementary Grades; The interaction of linguistic and arithmetic factors affects adult performance on arithmetic word problems

What should you do on the next reasoning-test word problem?

On the next word problem, name the quantity being asked for. State how the known amounts relate, briefly. Choose the smallest representation that clarifies that relationship, calculate, then put the answer back into the story: does it answer the question, use the right units, and preserve the stated relationship? This makes relation-first modeling more dependable than choosing an operation from a cue word alone. The unknown’s position determines which calculation fits.

If the answer is wrong, locate the failed step first. Did you misread the question, misunderstand a phrase, represent the relationship incorrectly, select an unsuitable operation, or make an arithmetic slip? Practice that operation with a fresh example, then check the reasoning instead of memorizing the previous item. Research on elementary instruction supports deliberate work with word-problem relationships and useful representations in those school settings; it does not show that this routine raises adult IQ-test scores or general intelligence.

For low-stakes practice with original reasoning questions, take Test IQ Free’s 50-question test and use its worked answers to review solutions. Its estimated IQ range is a provisional result from an unvalidated conversion with no validated adult reference population, not a normed score or confidence interval. The optional $9 report adds explanation of performance; it does not validate the estimate.

Questions readers ask