How to read a GMAT Focus Problem Solving stem without…

GMAT Focus Problem Solving items are the longest-standing question family in the exam's quantitative section, and on the current GMAT Focus edition they still carry the bulk of the arithmetic, algebra, and word-problem weight that a 60+ scaled score demands. A Problem Solving prompt is, in plain terms, a multiple-choice question with five answer options and exactly one correct value, drawn from a pool of content that spans arithmetic, number properties, algebra, word problems, rate and work, geometry, and counting. The candidate who treats every Problem Solving item as a single uniform type usually runs out of time, because the family contains a small set of recognisable sub-types, and each sub-type asks for a different first move. What follows is a working method: how to read a stem, which family the question belongs to, where the trap answers live, and how the answer choices should shape the calculation path. The goal is not to memorise a hundred tricks but to install a five-step discipline that survives the moment the test goes faster than the practice set did.

What a GMAT Focus Problem Solving item actually asks of you

A Problem Solving item presents a complete piece of information, then closes with a question that requires a single numerical or algebraic value. Five answer options sit beneath the prompt, labelled A through E, and exactly one of them is correct. There is no partial credit, no 'none of the above' as a default, and no credit for the method used; the score rewards the right letter. The unwritten contract is that the test-writer must construct five plausible options, which means the wrong four answers are themselves a piece of information. The classic move is to ignore the choices, compute the value, then scan; the smarter move is to read the choices early, because their shape tells you whether the question is testing an integer, a fraction, a sum of two terms, or a ratio.

Three features separate a GMAT Focus Problem Solving item from a school-level word problem. First, the question is always solvable in well under two minutes by a prepared test-taker, but only if the test-taker commits to a method before reaching for the calculator. Second, the wrong answers are not random; they are usually the result of a tempting shortcut that breaks at one step, such as cancelling before factoring, or treating a percentage increase as additive. Third, the answer choices are almost always ordered, either ascending or descending, which is a gift for the elimination method: once you know the answer must be larger than 165 and smaller than 175, you can rule out three options without finishing the calculation.

For most candidates, the first failure mode is reading the stem as a single block of text. The second is launching into arithmetic before deciding what the question is actually asking. The third is using the on-screen calculator on a step that the test-writer designed to be done by hand, which costs ten to fifteen seconds per item and compounds across a section. The remedy for all three is a structured read, and the read starts with the question sentence, not the story.

The five-step method that I install in every Problem Solving tutoring block

For most candidates, the difference between a 47 and a 60+ on Quant is not raw knowledge; it is the discipline of doing the same five steps on every item, in the same order, regardless of how the prompt feels. The method is short, and that is the point. It survives pressure because it does not require thought under pressure.

Step one is to read the last sentence first. The question almost always lives there. Reading it first turns the rest of the stem into a known-purpose data block instead of a story to interpret. Step two is to identify the family. The five recurring families are rate and work, mixture and weighted average, ratio and proportion, integer and divisibility, and algebraic expression. If you can name the family within five seconds, the rest of the method chooses itself. Step three is to set up variables on the scratch pad before computing. A two-letter table for rate problems, a single variable for algebraic expressions, and a fraction line for ratio questions are enough. Step four is to compute only what the question asks, no more. The trap answer that adds an extra step exists because the test-writer expected the candidate to compute one more line than was needed. Step five is to use the answer choices to back-solve when the direct path feels heavy, which is a feature of the format, not a compromise.

Why the first step is the last sentence, not the first

Most candidates read a Problem Solving stem top to bottom because that is how school teaches them to read. On the GMAT Focus this habit is expensive. The first two sentences of a word problem often contain background numbers that are not used; the question sentence carries the verb that defines the calculation. Reading the verb first reframes the entire block. A stem that begins with 'A certain mixture of acid and water' becomes a known-purpose data set the moment you know the verb is 'what fraction of the mixture is acid'. This is one of those tactical adjustments that adds back perhaps eight seconds per item, which across 21 items in a Quant section becomes the difference between finishing and bubbling in a guess at minute 40.

How to recognise the six recurring Problem Solving families

The pool of Problem Solving content is not infinite, and the test-writer draws from a small set of recognisable templates. Recognising the family within the first ten seconds is the single highest-leverage skill a candidate can install, because the family dictates the variable, the equation, and the trap. The six families below cover roughly 90% of the items you will meet.

Rate and work problems

Rate items describe machines, pipes, workers, or people completing a task at a stated rate, and they ask either for a combined rate, a completion time, or a comparison between two scenarios. The first move is to convert every rate into a single unit, usually 'per hour' or 'per minute', and then to add rates when work is in parallel, or to invert them when work is sequential. A common trap is the candidate who adds rates for a sequential task. A second trap is the 'start the second machine halfway through' variation, which forces a time-weighted split rather than a simple rate average. The answer choices usually include the value obtained by adding rates naively, which is the test-writer's way of penalising candidates who skip the variable step.

Mixture and weighted average problems

Mixture items present two solutions of different concentrations being combined, and they ask for the concentration of the mixture, or for one of the two original concentrations. The lever is the weighted average formula, which says the distance from the mixture to each original is inversely proportional to the volume of that original. The trap is to set up a single equation where the test-writer expected two unknowns; the cleanest version of the item is solvable by inspection once the weighted-average lever is in place. When the answer choices are far apart, the candidate can use the alligation method and avoid algebra altogether.

Ratio and proportion problems

Ratio items often hide a proportion inside a word problem. The reliable move is to assign a single variable to one part of the ratio and write the rest of the ratio in terms of that variable; the stem usually gives a sum or a product that closes the system. The trap is to assume the ratio is preserved after a change, when the change typically alters one term independently. For example, 'the ratio of boys to girls is 3 to 5, then 6 boys leave' produces a new ratio, and the candidate must track the change to the boys term only.

Integer and divisibility problems

Integer items ask for the largest value, the smallest count, or the remainder when a number is divided by another. The reliable move is to factor the divisor first, then test each answer choice for the condition. A second reliable move is to scan the answer choices for parity, because the test-writer will include at least one even answer when the correct answer is odd, and vice versa. The trap is to commit to long division on a divisor like 36 when the question only needs the remainder modulo 4.

Algebraic expression problems

Algebra items usually present a function, a relation between variables, and a question that asks for a specific value when a condition is met. The reliable move is to substitute simple values and compute, then re-substitute to confirm. The trap is to chase the algebraic expansion, which is the path the test-writer designed for candidates who do not see the substitution shortcut. For most candidates reading this, substitution is faster and more accurate than symbolic manipulation, and the only time to flip that default is when the substitution produces messy arithmetic.

Counting and probability problems

Counting items ask for the number of ways an event can occur, often with a constraint. The reliable move is to separate the constraint from the unconstrained count, then to subtract the forbidden cases. The trap is double-counting, which happens when the candidate applies two constraints independently. Probability items follow the same pattern: count the favourable, count the total, and divide, with care to avoid complementary-counting errors.

Reading the answer choices before you finish the calculation

The answer choices on a Problem Solving item are not decoration. They are an information channel, and the prepared test-taker reads them as soon as the family is identified. Three patterns show up often. The first is a wide spread, where the choices differ by tens or hundreds; in that case, a rough estimate is enough to eliminate three options, and the calculation only needs to land precisely on one of the remaining two. The second is a narrow spread, where the choices differ by less than 10% of the correct value; in that case, the calculation must be carried to full precision, and the on-screen calculator is justified. The third is a sum-of-two-terms pattern, where the choices are written as 'x + y' or as a two-digit number; in that case, the test-writer is signalling that the question rewards breaking the problem into two sub-computations rather than one large one.

Back-solving, which is plugging each answer choice into the stem and testing it, is a legitimate tactic on roughly one in five items. It is most useful when the family is algebraic and the equation is ugly. The cost is that four of the five choices will not satisfy the condition, so the candidate should expect to test two or three before finding the match. The benefit is that back-solving bypasses the algebra entirely. A reasonable rule is: if the direct path involves more than three lines of symbolic manipulation, switch to back-solving, and switch back as soon as the algebra becomes short enough.

Timing, pacing, and the minute-per-question budget

The GMAT Focus Quant section gives the candidate a fixed pool of minutes and a fixed pool of items. The minute-per-question budget is the only honest number to plan against, and it is tight. A reasonable target is two minutes per item, with a 30-second reserve for the last item or two. That budget is realistic, but only if the candidate does not over-invest in any single item. The practical rule is: if you are not moving toward a unique answer by minute two, mark the best available choice, flag the item, and move. A flagged item that you return to at minute 38 with a clearer head is worth more than an unfinished item that you rushed at minute 45.

Pacing also means accepting that not every item will yield to a clean method. Roughly 15% of items in a real Quant section are designed to feel hard, and the test-writer is not testing whether the candidate enjoys the item; the test-writer is testing whether the candidate will spend four minutes on an item that another candidate spent 90 seconds on and got right. The right response is to identify the hard item early, allocate the two-minute budget, and let it go. This is one of those adjustments that only feels right after several practice tests, because in the first weeks the candidate believes that more time on a hard item must mean a higher score. It does not. The score is a function of correct answers per minute, and a hard item returns less of that ratio than an easy item with the same letter.

Common pitfalls and how to avoid them on GMAT Focus Problem Solving

The first pitfall is reading the stem without reading the question. Candidates who read top to bottom often solve a problem the stem did not ask. The fix is mechanical: read the last sentence first, then read the stem with the question in mind. The second pitfall is over-reliance on the on-screen calculator. The calculator is allowed, but it costs time on every entry, and it encourages the candidate to compute more than the question requires. The fix is to do arithmetic by hand when the numbers are small and clean, and to reserve the calculator for steps where the arithmetic is genuinely heavy. The third pitfall is the trap answer that arises from cancelling before factoring, treating percentages as additive, or using the wrong base for a per-cent change. The fix is to write down the formula before plugging, especially on percentage items, where the formula choice decides the answer.

A fourth pitfall is the unit mismatch. A rate item will quote a rate in 'per hour' and ask for an answer in 'per minute', and the candidate who forgets to convert will land on a value that is one of the wrong choices. The fix is to circle the unit in the question sentence and the unit in the answer choices before starting the calculation. A fifth pitfall is the assumption that the answer choices are random; they are not. The wrong four answers are constructed from the same numbers as the right answer, which means each one corresponds to a single common error. A candidate who has the time can sometimes identify the trap by asking: 'which one of these answers would I get if I forgot to divide by two?' and then check whether that is the answer the test-writer placed in choice B.

How Problem Solving sits next to Data Sufficiency on the GMAT Focus

The GMAT Focus Quant section contains two item families, Problem Solving and Data Sufficiency, and the scoring treats them as a single pool. The candidate should know how the families differ, because the read of the stem is different. A Problem Solving item gives the data and asks for a value. A Data Sufficiency item gives the data in two statements and asks whether the two statements together are enough to determine the value. The trap is to answer the Data Sufficiency item as if it were a Problem Solving item, which is a common habit from earlier editions of the test.

Feature	Problem Solving	Data Sufficiency
Output asked	A single numerical or algebraic value	Whether the data is sufficient to determine a value
Number of answer choices	Five, A through E	Five, A through E, fixed wording
Common first move	Read the question, identify the family, set variables	Strip the data, restate the question, test statement 1 alone, then together
Typical trap	Computing one step more than the question asks	Solving for the value rather than judging sufficiency
Calculator use	Reserve for heavy arithmetic steps	Rarely useful; the question is structural
Best pacing signal	Two-minute budget per item	90-second budget per item

The two families share the same content pool, which means preparation for one trains the other. A candidate who drills rate items in Problem Solving is also building the recognition needed for rate items in Data Sufficiency. The two-minute versus 90-second budgets are the main difference, and they reflect that Data Sufficiency items are shorter to solve but easier to misread.

Building a Problem Solving drill plan that actually moves the score

A working drill plan has three layers. The first layer is recognition: 30 to 40 items per week, drawn from a single family, with the family identified before the calculation begins. The second layer is timing: 20 timed items per week, drawn from a mixed pool, with the two-minute budget enforced by a visible timer. The third layer is review: every item, right or wrong, revisited within 24 hours, with a one-line note on whether the failure was a recognition miss, a setup error, or a calculation slip. The reason for the third layer is that the score moves fastest when the candidate stops repeating the same error type, and a one-line note is enough to spot the pattern.

In my experience, candidates who skip the third layer plateau around a 47 to 51 on Quant, because their practice items are correct on average but the same two error types keep bleeding points. Candidates who do the third layer for at least six weeks typically climb into the 55 to 60+ band, because the error types get cleaned out one by one. The cost is roughly 90 minutes per week of review, which is less than a single additional timed section and is the single highest-leverage use of prep time on this section.

Pulling it together: a worked example of the five-step method

Consider a representative item: 'A tank is filled by pipe A in 6 hours and by pipe B in 4 hours. Pipe A runs alone for 2 hours, after which pipe B is also opened. How many additional hours does it take to fill the tank?' The last sentence is the question: how many additional hours. The family is rate and work. The variables are the rate of A, which is 1/6 tank per hour, and the rate of B, which is 1/4 tank per hour. The setup: in the first 2 hours, pipe A fills 2 × 1/6 = 1/3 of the tank. The remaining 2/3 of the tank is filled by both pipes, whose combined rate is 1/6 + 1/4 = 5/12 tank per hour. The time to fill 2/3 of the tank at 5/12 per hour is (2/3) ÷ (5/12) = (2/3) × (12/5) = 8/5 hours, which is 1 hour 36 minutes. The answer choices will include 1.6 hours, 1.6 as a fraction, and three traps: the candidate who forgets the first 2 hours will answer 8/5 from the start and pick a choice that does not match 1.6; the candidate who adds the rates naively will get a different number; the candidate who treats the question as 'how many hours total' will answer 3.6 hours. The correct answer is 1.6 hours, and the trap answers sit in the four other choices.

Notice what the method did. It read the last sentence first, named the family, set the variables, computed only what the question asked, and used the answer choices to confirm. The whole item took under 90 seconds. A candidate who read the stem top to bottom, jumped to the calculator, and computed the total time would have spent twice as long and landed on the wrong letter. The method is the score, not the arithmetic.

TestPrep Europe's diagnostic assessment is the natural starting point for candidates who want their Problem Solving error profile mapped before they commit to a drill plan.

Conclusion and next steps

GMAT Focus Problem Solving rewards a five-step method, a working knowledge of six recurring families, and a pacing discipline that treats the two-minute budget as a hard ceiling. The candidate who installs the method, drills the families in isolation, and reviews every item within 24 hours will move the score more in six weeks than in the previous six months of unfocused practice. The next concrete step is a single timed set of 20 mixed items, with the two-minute budget enforced and a review log kept for each one. From that log, the drill plan for the following week is set by the error type that appeared most often, and the cycle continues until the error types thin out and the score plateaus at the level the candidate can defend on test day.

Frequently asked questions

How many Problem Solving items appear on the GMAT Focus Quant section?

The Quant section mixes Problem Solving and Data Sufficiency items into a single 21-question pool, with Problem Solving usually making up the larger share. The exact split varies between forms, so a candidate should prepare for both families rather than banking on a specific count.

Is the on-screen calculator allowed on every Problem Solving item?

Yes, the calculator is available throughout the Quant section, but the items are designed so that heavy calculator use is a pace penalty. Use it for genuinely heavy arithmetic and do small, clean steps by hand.

What is the single highest-leverage habit for raising a Problem Solving score?

Reading the question sentence before the rest of the stem, then identifying the item family within the first ten seconds. This single habit eliminates the most common error type, which is solving a problem the question did not ask.

How long should a candidate spend on a single Problem Solving item?

Target two minutes per item, with a 30-second reserve across the section. If a clear path is not visible by minute two, mark the best available choice, flag the item, and return to it only if time allows at the end of the section.

Does Problem Solving preparation carry over to Data Sufficiency on the GMAT Focus?

Yes. The two families share the same content pool, so drilling rate, mixture, ratio, and integer items in Problem Solving also builds the recognition needed for Data Sufficiency. The main difference is the read of the stem and a slightly tighter time budget on Data Sufficiency items.

How to read a GMAT Focus Problem Solving stem without losing the first 20 seconds