How does GMAT Focus Two-Part Analysis actually score your…

Two-Part Analysis is one of five item families inside the Data Insights section of the GMAT Focus, and it is the only one that forces candidates to solve two related problems in a single stem. A single item carries a stem with a short business or quantitative scenario, two open blanks or two-part question markers, and a list of answer choices that pair one answer from the first part with one answer from the second. The GMAT Focus exam format presents 20 Data Insights questions in a 45-minute window, and Two-Part Analysis typically accounts for a meaningful slice of that count, so preparation strategy must treat it as a recurring question type rather than a curiosity.

The score you earn on each Two-Part Analysis item is binary at the response level: the GMAT Focus scoring engine awards the point only when both selected options are correct, and awards nothing when either is wrong. This single design choice has enormous consequences for how a candidate should study, practise, and time-budget the section. The rest of this article breaks the prompt apart, names the recurring question types, gives a working method for the first 60 seconds of an item, and flags the pitfalls that drain the most points in real test conditions.

How Two-Part Analysis fits inside the GMAT Focus Data Insights section

Data Insights is the third and newest section of the GMAT Focus, and it is the section that most reshapes preparation strategy relative to the legacy GMAT exam. The section blends quantitative reasoning, verbal reasoning, and data literacy into a single 45-minute module containing 20 scored questions drawn from five item families: Data Sufficiency, Multi-Source Reasoning, Table Analysis, Graphics Interpretation, and Two-Part Analysis. The exact count of each family varies across the adaptive form, but in published official practice materials Two-Part Analysis tends to appear between two and four times per section. That range is enough to make a candidate's Data Insights score sensitive to it, especially because each Two-Part Analysis item is unusually heavy in cognitive load.

The format itself is distinctive. The candidate reads a short scenario — often a few sentences framing a business problem, a trade-off, or a quantitative relationship. Below the stem, the prompt identifies a first decision and a second decision, each phrased as its own open-ended question. To the right of the stem, a list of answer options pairs a value or statement for the first decision with a value or statement for the second. Candidates select one answer for each part by clicking the row that contains the correct pair. The pair is graded as a single unit: one credit, zero credit, no partial credit. The exam format therefore looks superficially like a standard multiple-choice item, but the cognitive demand is closer to two parallel mini-problems fused into one prompt.

For preparation strategy, the practical consequence is that Two-Part Analysis consumes more working memory than its question count suggests. A candidate who treats it as 'just another Data Insights question type' will under-budget mental energy and arrive at the second half of the section fatigued. A candidate who treats it as a distinct work-unit with its own first-60-seconds protocol tends to perform more consistently. The remainder of this article focuses on that protocol, the prompt shapes the test reuses, and the scoring logic that drives the item family.

The anatomy of a Two-Part Analysis prompt

Every Two-Part Analysis item has three mechanical parts: a stem, two parallel sub-questions, and a paired-options list. The stem is usually 40 to 90 words long and describes a real or near-real situation, such as allocating a budget between two projects, choosing between two suppliers with different cost structures, or comparing two rate-sensitive loan scenarios. The two sub-questions are nearly always phrased as open questions, meaning they do not have a fixed set of choices — instead, the candidate is asked to find a value, a statement, or an option that satisfies the condition. The paired-options list then offers between four and six rows, each row containing one candidate for the first blank and one candidate for the second blank.

The mechanical difference from a standard multiple-choice question is worth dwelling on. In Data Sufficiency, the candidate is asked a single question and selects one of five fixed statements. In Multi-Source Reasoning, the candidate reads a short dossier and answers several questions about it. In Two-Part Analysis, the candidate is asked to make two decisions simultaneously, and the answer options are organised as pairs. This pairing has an important side-effect: it is sometimes possible to determine one of the two sub-answers confidently while remaining uncertain about the other. The exam format forbids selecting a confident answer for part one and a tentative answer for part two from different rows — the selection is a single row, scored as a single unit.

The pairing mechanic is also why the prompt shapes the test reuses are so recognisable. Once a candidate has seen three or four Two-Part Analysis items in practice, the structural variety collapses to about five recurring shapes, and recognition itself becomes preparation. Recognising the shape in the first 30 seconds of an item buys back 30 to 60 seconds of working time, and over a 45-minute section that compounding effect is the difference between finishing strong and finishing ragged.

What the stem is actually doing

The stem is doing one of two jobs, and naming which one is the first tactical move. About half the prompts are numerical Two-Part Analysis items: the candidate must compute a value for each of two entities, usually under a shared constraint. The other half are logical items: the candidate must decide which of two qualitative statements is true, often by evaluating a chain of conditional relationships. Numerical items tend to centre on rate × time, weighted averages, profit margin, or break-even logic. Logical items tend to centre on 'if-then' chains, mutually exclusive categories, or a single categorical assignment that drives both sub-questions.

A useful first-60-seconds protocol is to identify which type you are facing before reading the answer options. Numerical stems usually contain explicit figures — a table of unit costs, two interest rates, a fixed budget — and a request to compute two values. Logical stems usually contain named entities (a project, a supplier, a candidate) and a request to assign one of several categorical options to each. The skill here is the willingness to spend five seconds of silence classifying the item before attacking the math or the logic. Most candidates read straight from top to bottom and only later discover that they solved the wrong question.

How GMAT Focus scoring treats the paired response

On the GMAT Focus scoring report, Two-Part Analysis is reported as part of the Data Insights scaled score, which runs on a 60-to-90 band, and each item is treated as a discrete unit of credit. The candidate's raw score on the section is the number of items answered correctly, converted to the scaled score through an equating process that depends on item difficulty and adaptive routing. The crucial detail is that the equating is applied at the item level, not the sub-question level. Getting part one right and part two wrong yields zero credit on the item. There is no partial credit. The Data Insights section therefore treats every Two-Part Analysis item as an all-or-nothing bet, and preparation strategy must respect that asymmetry.

The equating logic also explains why a careless error on a Two-Part Analysis item is more costly than a careless error on a single-answer Data Sufficiency item. On Data Sufficiency, the candidate can miss a step inside the reasoning and still pick the correct statement; the partial-thinking penalty is lower. On Two-Part Analysis, any error in either sub-question voids the entire item, so the candidate's working accuracy has to be at least as high as on a quantitative problem with a single answer. In practice this means the section's effective difficulty curve for Two-Part Analysis is steeper than its position on the form suggests.

There is one more scoring subtlety worth knowing. The official score report breaks the Data Insights performance into a small number of skill bands rather than per-item, so the candidate does not see 'Two-Part Analysis: 3 of 4 correct'. The performance feedback shows aggregate reasoning skills — for example, 'analyzing quantitative relationships' or 'solving multi-step problems' — and the candidate has to interpret whether Two-Part Analysis is a relative weakness. For preparation strategy, this is a signal to keep an item-by-item error log rather than rely on the score report alone, because the report does not separate the two parts of a Two-Part Analysis question for credit attribution.

Five prompt shapes the GMAT Focus reuses across forms

Across the published official practice materials, Two-Part Analysis items cluster into five recognisable shapes. Naming them in advance is preparation, not prediction, and the candidate should expect to see at least one of the first three in any given form.

The paired arithmetic shape: a single constraint governs two entities, and the candidate computes a value for each. Classic example: a fixed budget split between two advertising channels with different cost-per-impression rates, where the first sub-question asks for the dollar amount in channel A and the second asks for the dollar amount in channel B. The numerical work is light; the test is on whether the candidate set up the constraint correctly.
The break-even shape: two products or services have different cost and revenue structures, and the prompt asks for the quantity at which total cost equals total revenue for each. The pair is symmetric, and the trap is solving both with the same formula when the structures are actually different.
The weighted allocation shape: a weighted average is given for a portfolio of two assets, and the candidate must find the weight of one asset and the weight of the other. The trap is reversing which weight belongs to which asset, which silently produces a wrong pair even when the math is right.
The categorical assignment shape: a small set of named entities (projects, candidates, suppliers) must each be assigned a category (priority level, region, qualification), and the candidate must pick the assignment that satisfies two stated rules. This is the most verbal of the five shapes and the one where English reading discipline matters most.
The conditional chain shape: a series of 'if-then' statements connects several variables, and the candidate must select two values that are both consistent with a stated conclusion. The trap is treating the chain as a list of independent facts rather than a directed dependency.

Most candidates encounter two of the first three shapes and one of the last two in a typical Data Insights form. The recognition payoff is real: once the shape is named, the candidate knows the relevant formulas or the relevant logical structure, and the item compresses from a 2-minute problem to a 75-second one.

A first-60-seconds protocol that holds across all five shapes

The single biggest gain in Two-Part Analysis preparation is a protocol that fits inside the first minute of an item. After working with several hundred of these items, I have settled on a four-step opening that most candidates can internalise in two weeks of daily practice. None of the four steps requires the answer options, which is the point — they consume the stem only and protect the candidate from prematurely anchoring on a paired answer.

Step 1, classify the shape. Read the stem and decide whether it is numerical or logical, and which of the five prompt shapes it most closely resembles. Five seconds. If you cannot classify, you are not ready to read the options yet.

Step 2, name the unknown for each part. The stem usually asks the same kind of question in two places: 'How many units of A?' and 'How many units of B?', or 'Which supplier for project 1?' and 'Which supplier for project 2?'. Write the two unknowns in shorthand. This is the single most skipped step, and skipping it is the single most common cause of a part-swap error.

Step 3, write the constraint. The constraint is the equation or rule that links the two parts. For a paired arithmetic item, it is usually a sum, a weighted average, or a fixed total. For a categorical assignment, it is the second rule in the prompt. Putting the constraint into a single sentence forces clarity.

Step 4, scan the options for the cheapest filter. Now — and only now — look at the paired options. Apply the cheapest filter first. If part one is a numeric value, eliminate rows whose first entry cannot match the part-one calculation. If part one is a categorical label, eliminate rows whose first entry contradicts a named rule. The second filter is applied to the survivors. The first-60-seconds protocol ends here, and the candidate moves into the main calculation.

This protocol is deliberately not a content strategy. It is a working-memory strategy. Its purpose is to keep the candidate from solving the right math in the wrong order, which is the dominant failure mode on Two-Part Analysis items under time pressure.

Common pitfalls and how to avoid them

Two-Part Analysis is a question type where a small number of recurring errors account for most lost points, and the errors are unusually easy to repeat because the prompt structure makes them feel correct. Below are the four most common pitfalls, with the specific defensive move for each.

Part-swap error: the candidate solves correctly for both unknowns but pairs them in the wrong row. Defensive move: always write the two unknowns in the same shorthand as the options. If the first option column is labelled A and the second is labelled B, your shorthand must be in the same order. A surprisingly high share of the errors my students log on this question type are pure label swaps.
Single-constraint overreach: the candidate uses only one of the two constraints in the stem and produces a pair that satisfies the first constraint but not the second. Defensive move: after finding a candidate pair, mentally re-check it against both constraints before locking the row. The exam format does not flag a partial fit, so the candidate must run the re-check themselves.
Premature anchoring on the first row that 'looks right': Two-Part Analysis options are arranged as a small table, and a pair can look visually plausible even when it is wrong. Defensive move: never lock a row based on a single part being correct. The scoring is all-or-nothing, so a single confident part is not a stopping condition.
Time bleed on a single hard item: the candidate spends three minutes on a Two-Part Analysis item and arrives at the answer with high confidence, then runs out of time on later Data Insights items that they would have got right at normal pace. Defensive move: cap each Two-Part Analysis item at roughly 2.5 minutes in the timed section, and flag any item over 3 minutes as a future decision to guess-and-move.

There is a fifth pitfall that is not strictly an error but a strategic mistake: avoiding Two-Part Analysis in practice. Some candidates treat the item family as too hard and refuse to drill it, which depresses the Data Insights score ceiling because the adaptive form keeps re-presenting similar shapes. Preparation strategy for Two-Part Analysis must include deliberate, timed exposure, not avoidance.

How Two-Part Analysis interacts with the adaptive logic of the GMAT Focus

The GMAT Focus is a multi-stage adaptive exam: each section begins with an unscored preview, then a routing module whose difficulty is held constant, then a second module whose difficulty is calibrated by performance on the first. Data Insights is no exception, and the routing inside the section means that a strong performance on the first module tends to deliver a more difficult second module. Two-Part Analysis items appear in both modules, and the difficulty of the items you face is sensitive to your performance on the earlier items in the section — including other Two-Part Analysis items.

The implication for preparation strategy is that Two-Part Analysis is not a stand-alone skill. A candidate who has drilled Two-Part Analysis in isolation but underperforms on Multi-Source Reasoning in the first module will see a lower-difficulty second module, where the Two-Part Analysis items themselves are easier. Conversely, a strong first module lifts the difficulty of the second module, and the Two-Part Analysis items in that second module are more likely to be the harder prompt shapes. Performance on the item family is therefore co-dependent on the rest of the section, and the section-level pacing plan has to budget for that.

From a question-types perspective, the harder prompt shapes tend to be the conditional chain and the categorical assignment, because the reading load is heavier and the elimination work is more delicate. The easier shapes tend to be the paired arithmetic and the break-even, because the math is more constrained. A well-built preparation plan should treat both ends of that spectrum as drill targets, not just the harder end, because the easier shapes are the ones that build the routing score that unlocks the higher-difficulty second module.

Building a Two-Part Analysis preparation plan that fits the GMAT Focus

Effective preparation for Two-Part Analysis in the context of the full GMAT Focus exam is a balance between item-family-specific drilling and section-level pacing. Below is a six-week framework that has worked for most of the candidates I have tutored, calibrated to roughly 10 to 12 hours of focused study per week. The numbers are practical targets, not promises, and they should be adjusted for the candidate's starting score and target score.

Week 1: shape recognition and stem classification

The first week is about pattern recognition, not problem solving. Work through 30 to 40 Two-Part Analysis items at untimed pace, with the answer choices covered. For each stem, name the shape from the five above, classify it as numerical or logical, and write the two unknowns before looking at the options. The goal is to make the classification automatic by the end of the week, so the first 30 seconds of every future item are spent on shape, not on re-reading.

Week 2: first-60-seconds protocol under light time pressure

In the second week, repeat the same item pool under a soft time cap of roughly 2 minutes per item. The cap is deliberately generous, because the goal is to internalise the protocol, not to chase a speed number. Keep an error log that distinguishes part-swap errors, single-constraint overreach, and pure knowledge gaps. Most candidates discover that one of those three categories dominates, and the error log tells them which defensive move to drill.

Week 3: section-level integration

The third week moves out of isolated Two-Part Analysis practice and into mixed Data Insights sets. Pull 20-question Data Insights sections from official practice materials and work them in 45-minute timed conditions. The aim is to see how the item family interacts with the other four. Most candidates find that Two-Part Analysis items feel heavier than the others, and the section-level pacing must absorb that weight.

Weeks 4 to 6: adaptive-condition simulations and review

The final three weeks are spent in adaptive-condition full-length simulations, with deliberate over-learning on the prompt shapes that the candidate's error log flagged. A useful pattern is to spend one session per week on a single weak shape, drilling 12 to 15 items in a tight loop until the failure rate drops below roughly 20 per cent, before returning to mixed practice. The scoring report's skill bands should start to reflect movement by the end of week 6 for most candidates.

Calibrating the time budget

A useful rule of thumb for section-level pacing: budget around 2 to 2.5 minutes per Two-Part Analysis item in timed conditions, and 1.5 to 2 minutes per item on the lighter item families. Across 20 Data Insights items, that budget fits comfortably inside the 45-minute window, with a 2 to 3 minute reserve for the final review pass. The reserve matters more on Data Insights than on the other two sections because the paired-response scoring means a careless last-minute error is more expensive.

Reading the official materials for the patterns that actually repeat

Official GMAT Focus practice materials are the single most cost-effective preparation resource for Two-Part Analysis, and most candidates underuse them. The right way to read them is to focus on three layers: the stem, the constraint, and the option table. Treat the stem as a small case study rather than as a problem to solve, and ask why the test chose the figures it chose. Most items are built around a single numerical relationship, and the test-writer is signalling that relationship by including or omitting specific data points. Candidates who learn to read the omission as a clue tend to find the constraint faster than candidates who treat every number as load-bearing.

The option table is the second underused layer. The pairing is not random: usually three of the rows share a common first-part value or a common second-part value, and the one row that differs is the only viable candidate. Recognising that the test-writer is offering three near-misses and one correct pair is preparation, and it short-circuits a lot of unnecessary calculation. The skill is to read the table as a designed object, not as a list of guesses.

Putting the pieces together on test day

On test day, the candidate who has internalised the first-60-seconds protocol, the five prompt shapes, the part-swap defensive move, and the section-level time budget will experience Two-Part Analysis as a familiar work-unit rather than a surprise. The item is still demanding, but it is no longer the only question type in the section that demands careful reading. The skill that ties everything together is the willingness to spend the first 30 seconds of an item in silence, classifying and naming, before touching the answer options. In my experience this single change accounts for more of the score lift on Two-Part Analysis than any content review, because the items are not really testing content; they are testing structured reading under paired-response pressure.

Conclusion / Next steps. The path forward is to anchor the next two weeks of study on shape recognition and the first-60-seconds protocol, using a tight error log to surface the dominant failure mode, and only then escalate into section-level timed practice. TestPrep Europe's Two-Part Analysis diagnostic is a natural starting point for candidates who want a structured first pass through the five prompt shapes with a tutor's review of their part-swap and single-constraint failure rates.

FAQ

Frequently asked questions

How many Two-Part Analysis items appear on a typical GMAT Focus form?

The exact count varies across the adaptive forms, but published official practice materials show Two-Part Analysis appearing between two and four times inside the 20-question Data Insights section. The shape distribution tends to include the paired arithmetic and break-even patterns at minimum, with the weighted allocation, categorical assignment, and conditional chain shapes appearing depending on the routing module's difficulty.

Is partial credit ever possible on a Two-Part Analysis item?

No. GMAT Focus scoring treats each Two-Part Analysis item as a single all-or-nothing unit. The candidate selects a row from the paired options list, and the row is graded as one decision. Getting the first part correct and the second part wrong yields zero credit on the item, which is why the first-60-seconds protocol insists on naming both unknowns before reading the options.

How much time should a candidate spend on a single Two-Part Analysis item?

In timed section conditions, around 2 to 2.5 minutes is a reasonable working budget per item. Items over 3 minutes should be flagged as guess-and-move candidates, because the Data Insights section is only 45 minutes long and 20 questions overall, so time bleed on a single item directly threatens the rest of the section's pacing.

Does Two-Part Analysis appear in both adaptive modules of Data Insights?

Yes. Two-Part Analysis items can appear in either the first or the second module, and the difficulty of the items is sensitive to the candidate's performance on the earlier items in the section. Strong performance on the first module tends to deliver a higher-difficulty second module, which raises the likelihood of the harder categorical assignment and conditional chain shapes.

What is the most common error on Two-Part Analysis in real test conditions?

Part-swap errors are the most common silent failure mode: the candidate solves both sub-questions correctly but pairs the answers in the wrong row because the shorthand for the two unknowns was reversed relative to the option columns. The defensive move is to label the two unknowns in the same order as the options before starting the calculation, and to re-check the pairing before locking the row.

How does GMAT Focus Two-Part Analysis actually score your reasoning?