GMAT two-part analysis questions are one of five question types in the Data Insights section, and they ask you to pick one answer in Column A and another in Column B from a shared list of choices. Because there is no partial credit, a disciplined strategy matters more here than on most other GMAT question types. This guide breaks down the format, shows you when you will see these questions, walks through a six-step solving framework with a worked example, and highlights the mistakes that cost students the most points on test day.
A GMAT two-part analysis question starts with a short prompt that could be quantitative, verbal, or logic-based. Below the prompt you will see a three-column answer table: one column lists five or six answer choices, and the other two columns, labeled Column A and Column B, each require one selection. Your job is to pick exactly one option per column so the pair together satisfies the prompt.
The format is unique to two-part analysis questions and is what makes them so distinctive in the Data Insights section. Column A asks one thing (say, the total number of shares purchased), Column B asks another (the benefit per employee), and the answer choices column provides the shared pool. Both selections can come from the same row of the list or different rows — there is no rule forcing them apart.
This is the single most important fact about two-part analysis questions. If you answer Column A correctly but miss Column B, you score zero on the question — exactly the same as missing both. That zero-tolerance rule is why a framework that links the two columns is more valuable than solving each one separately.
Two-part analysis is one of five question types in the GMAT Focus Data Insights section. The others are Data Sufficiency, Multi-Source Reasoning, Table Analysis, and Graphics Interpretation. All five types share the same no-partial-credit rule and draw on the same underlying reasoning skills, so time you spend drilling TPA pays off across the whole section.
| Attribute | Value |
|---|---|
| Section | Data Insights |
| Questions per exam | 2-4 (10-20% of Data Insights) |
| Answer format | Pick one choice per column (Column A and Column B) |
| Answer choices | 5 or 6 per question |
| Possible answer pairs | 25-36 combinations |
| Target time per question | 2.5-3 minutes |
| Hard time cap | 4 minutes |
| Partial credit | None - both parts must be correct |
GMAT TPA questions are a small slice of a tight section, so knowing their share helps you calibrate both prep time and pacing on test day. The GMAT Focus Data Insights section contains 20 questions to be completed in 45 minutes, and two-part analysis makes up 10-20% of those questions. That works out to roughly 2-4 TPA questions per exam.
The 20 questions are distributed across the five question types with rough percentage ranges set by GMAC. Data Sufficiency has the largest share (20-40%), Graphics Interpretation is next (20-30%), and Multi-Source Reasoning, Table Analysis, and Two-Part Analysis each fall in the 10-20% band. Any given test lands somewhere inside those ranges because the section is computer-adaptive.
At 10-20%, you should plan for between 2 and 4 GMAT Data Insights two-part analysis questions on your exam. If you hit the per-question target of 2.5-3 minutes, those 2-4 questions consume roughly 5 to 12 of your 45 Data Insights minutes — meaningful enough that a bad TPA pacing run can derail the rest of the section.
The section is computer-adaptive, which means the test engine chooses the next question based on your performance so far. Difficulty and question-type mix shift slightly from test to test, but the 10-20% range is a published constraint — you will not see zero TPA questions on one exam and ten on another.
Every GMAT two-part analysis question you see belongs to one of three subtypes. Classifying the prompt within the first 15 seconds determines which mental toolkit to pull out: an equation for quantitative, premise-and-conclusion reasoning for verbal, or a rules diagram for logic-based.
| Subtype | What It Tests | Telltale Signs | Best Strategy |
|---|---|---|---|
| Quantitative | Algebra, rates, ratios, percentages, word problems | Numbers, variables, or formulas in the prompt | Write the governing equation, then eliminate by constraint |
| Verbal / Critical Reasoning | Strengthen, weaken, inference, assumption | Short argument-style passage, no math | Use CR strategies; identify premise vs conclusion |
| Logic-based | Applying explicit rules to a scenario | Bulleted rules, scheduling, or assignment puzzle | Sketch a quick diagram and test rules one at a time |
The good news is that the underlying skills are identical to GMAT Problem Solving and Critical Reasoning. The novelty of TPA is the two-answer format, not the content. That means your existing prep for Quant and Verbal transfers directly — you are mainly learning a new way to deliver answers you can already find.
With 5-6 answer choices per column, a typical TPA question offers 25-36 possible answer pair combinations. Brute force across that space is a losing game. A repeatable six-step process keeps your work focused, eliminates bad pairs early, and leaves you with a small set of candidates to actually test.
Step 1 — Classify the subtype. Scan the prompt and decide: quantitative, verbal, or logic-based. This takes 10-15 seconds but determines everything that follows.
Step 2 — Read each column header separately. Column A and Column B ask different things. Underline the key noun phrase in each header, especially any min/max language or "must be" versus "could be" distinctions.
Step 3 — Write the link. On scratch paper, capture the relationship that ties Column A to Column B. For quantitative, that is usually an equation. For verbal, it might be "Column A strengthens; Column B weakens the same conclusion." For logic-based, it is the rule chain that connects both answers to the scenario.
Step 4 — Eliminate by constraint. Go through the answer choices and cross off anything that violates a stated constraint before you do any heavy solving.
Step 5 — Solve or test the survivors. With the field narrowed, either solve directly or plug in the few remaining candidates.
Step 6 — Verify the pair together. Before you click confirm, check that the final pair satisfies every constraint as a unit — not just each column individually.
Worked Example — Quantitative TPA
A corporation with N total employees has R percent of employees enrolled in a stock option plan, receiving S shares each year at an average price of P dollars per share.
Column A: the total number of shares the corporation must buy annually.
Column B: the average monetary value of a single enrolled employee's annual benefit.
Answer choices: (A) N*R (B) N*S (C) NRS/100 (D) RP/100 (E) S*P (F) RSP/100
TPA questions earn a slightly longer time budget than most other Data Insights question types because you are effectively solving two linked sub-problems. The tradeoff is that TPA can blow up your pacing faster than any other DI type if you let it.
Aim for 2.5 to 3 minutes per two-part analysis question. Treat 4 minutes as a hard cap — if you cross it, eliminate aggressively, pick the best pair, and move on. The Data Insights section averages about 2.25 minutes per question across all five types, so giving TPA 3 minutes means borrowing a little time from the easier question types like Data Sufficiency.
The table below shows how TPA pacing fits into the full 45-minute section. Use it as a budget, not a rule — if you are faster on Data Sufficiency than the target, you can afford a slightly longer TPA; if you are slower, tighten up.
| DI Question Type | Approx. Share of Section | Target Time | Hard Cap |
|---|---|---|---|
| Data Sufficiency | 20-40% | 1.5-2 min | 3 min |
| Graphics Interpretation | 20-30% | 2-2.5 min | 3.5 min |
| Table Analysis | 10-20% | 2-2.5 min | 3.5 min |
| Two-Part Analysis | 10-20% | 2.5-3 min | 4 min |
| Multi-Source Reasoning | 10-20% | 2-3 min | 4 min |
| Section average | 100% / 20 Qs in 45 min | 2.25 min | N/A |
Going over 4 minutes on one TPA often costs 2-3 easier questions later in the section. When you hit the cap, eliminate every answer choice you can rule out with a constraint, commit to the most-supported pair from what remains, lock it in, and move on. Protecting the rest of the section is always worth more than grinding out one stubborn question.
Estimate how much of your 45-minute Data Insights section two-part analysis will consume, and how much time is left for the other question types.
Most missed two-part analysis questions are process errors, not content gaps. Students already know the algebra or the Critical Reasoning skill — they just apply it wrong in the two-column format. These are the top GMAT two-part analysis tips derived from audit logs and tutor feedback on missed questions.
The number-one process error is solving each column in isolation instead of as a linked pair. A close second is assuming the two answers must come from different rows — there is no such rule, and valid same-row pairs are often the designed answer.
Misreading "minimum" vs "maximum" in a column header flips the answer entirely. Ignoring an explicit constraint stated in the prompt forces you to rework the problem when the final pair fails verification. Underline the superlative in each header before solving.
Random plug-and-play across 25-36 answer combinations burns minutes fast. And blowing past the 4-minute cap on a hard TPA often means 2-3 easier questions later in the section get rushed and missed.
| Mistake | Why It Hurts | Fix |
|---|---|---|
| Solving each column alone | Misses the linkage; you get one column right and the other wrong | Write the relationship between columns before solving |
| Assuming different rows | Eliminates valid same-row pairs | Consider every pair, including same-row |
| Misreading min vs max | One wrong word flips the answer | Underline the superlative in each column header |
| Random plug-and-play | Wastes time across 25-36 pair combinations | Eliminate by constraint before testing pairs |
| Ignoring constraints | Forces rework when the final pair fails | List every explicit constraint on scratch paper |
| Going past 4 minutes | Costs 2-3 easier DI questions later | Set a mental cap; guess and move if you hit it |
Effective GMAT two-part analysis practice prioritizes framework-building and error-log review before it prioritizes volume. Running 50 questions a week with no reflection will not move your TPA accuracy — running 15 questions and reviewing every miss carefully will.
Start with untimed practice. Work a question, apply every step of the six-step framework explicitly on paper, and only then check the answer. The goal is to make the framework automatic before time pressure is added. Once your accuracy is stable above about 70%, layer the timer on.
The GMAT Official Guide and the GMAT Focus Official Practice Exams are the best TPA sources because they are written by the same team that writes the real test. Supplement with top third-party materials, but treat official questions as the benchmark. Mix TPA drills with the other four Data Insights question types in every session to rehearse real section flow.
Keep a simple error log. For each missed TPA question, write down whether the error was a classification problem (wrong subtype), a setup problem (wrong equation or missed constraint), or an arithmetic problem (right setup, wrong calculation). Review the log at the end of every week and attack the most frequent error type first.
Test the six-step framework on five short GMAT two-part analysis practice questions covering format, timing, classification, and process mistakes. Commit to an answer before clicking Check.
Two-part analysis questions make up 10 to 20 percent of the Data Insights section, which contains 20 questions total. That works out to roughly 2 to 4 two-part analysis questions on any given GMAT Focus test. Because the section is computer-adaptive, the exact count and difficulty vary from test to test, but the percentage range is fixed by the test maker.
Most GMAT tutors recommend a target of 2.5 to 3 minutes per two-part analysis question, with a hard maximum of 4 minutes. The Data Insights section averages about 2.25 minutes per question across all five question types, so TPA gets a slightly larger budget. If you pass the 4-minute mark, eliminate what you can, pick the best pair, and move on to protect the rest of the section.
No. You must select the correct answer in both columns to earn credit on a two-part analysis question. Getting one column right and the other wrong scores the same as missing both. This is why a strategy that links the two columns together is essential rather than solving each column independently. The same no-partial-credit rule applies across the entire Data Insights section.
Yes. Nothing in the two-part analysis format forces the two selected answers to be in different rows of the answer table. A common trap is assuming Column A and Column B must come from separate rows, which leads students to eliminate valid pairs. Always consider every possible pairing, including same-row combinations, when checking which pair satisfies the prompt's constraints.
They are not inherently harder, but they tend to take longer. Data Sufficiency questions typically take 1.5 to 2 minutes each, while two-part analysis usually takes 2.5 to 3 minutes because you are effectively solving two linked sub-problems. Many students find TPA less tricky than Data Sufficiency because the question is more explicit, but the time cost makes pacing more critical.
Two-part analysis questions test the same skills as GMAT Problem Solving and Critical Reasoning, applied in a linked two-column format. Quantitative TPA tests algebra, word problems, rates, and ratios. Verbal TPA tests inference, strengthen and weaken reasoning, and passage comprehension. Logic-based TPA tests constraint application. The novelty is the two-answer format, not the underlying content.