GMAT Data Insights Section 2026

The Complete Breakdown of the Most Underestimated Section on the GMAT Focus Edition

---

Most GMAT candidates walk into their first practice test treating Data Insights as an afterthought. They spend months drilling Quant and Verbal, then allocate the final two weeks to "that other section." The results are predictable: a Data Insights score that drags down an otherwise competitive performance, often by more points than any single Quant or Verbal weakness.

Here is why that mistake is so costly in 2026: Data Insights is not a supplementary section. It accounts for one-third of your total GMAT Focus score. A weak Data Insights performance does not just hurt your score — it caps it. No amount of Quant excellence compensates for a 60th-percentile Data Insights result when all three sections contribute equally to your final score.

This guide gives you everything you need to understand, prepare for, and excel at Data Insights. We cover all five question types in depth, break down the time allocation challenge, and give you a targeted preparation framework to make Data Insights a strength — not a liability.

---

Section 1: What Is the GMAT Data Insights Section?

Data Insights replaced the old Integrated Reasoning section and absorbed the Data Sufficiency question type from the classic GMAT's Quantitative section. The result is a section that tests something genuinely distinct from either Quant or Verbal: your ability to synthesize information from multiple formats, evaluate whether data is sufficient to answer a question, and reason clearly under time pressure when the inputs are messy or complex.

The structure at a glance:

Attribute	Detail
Number of questions	20
Time allowed	45 minutes
Average time per question	~2 minutes 15 seconds
Contribution to total score	One-third of GMAT Focus score
Question types	5 distinct types
Calculator available	Yes — on-screen calculator

The on-screen calculator is available throughout Data Insights — a meaningful difference from the Quant section. However, do not let this lull you into a false sense of security. The section is not primarily about arithmetic. It is about reasoning, interpretation, and judgment. The calculator helps with computation; it cannot help you decide what to compute or whether the data in front of you actually answers the question being asked.

---

Section 2: The Five Question Types — In Depth

Understanding each question type is not optional preparation — it is the foundation. Candidates who approach Data Insights without knowing what each type looks like, how it is scored, and what cognitive traps it sets are essentially solving puzzles without knowing the rules.

Here is a complete breakdown of all five types, weighted by their approximate share of the 20-question section.

---

1. Data Sufficiency (DS) — ~30% | ~6 Questions

Data Sufficiency is the most heavily weighted question type in Data Insights and arguably the most conceptually unusual question format on any standardized test. It asks you not to solve a problem, but to evaluate whether you could solve it given the information provided.

The format: Every DS question presents a question stem followed by two statements, labeled (1) and (2). Your task is to determine whether Statement 1 alone is sufficient to answer the question, Statement 2 alone is sufficient, both together are sufficient, either alone is sufficient, or neither together is sufficient.

The five answer choices are always the same:

• (A) Statement (1) alone is sufficient, but statement (2) alone is not sufficient
• (B) Statement (2) alone is sufficient, but statement (1) alone is not sufficient
• (C) Both statements together are sufficient, but neither alone is sufficient
• (D) Each statement alone is sufficient
• (E) Statements (1) and (2) together are not sufficient

Because the answer choices never change, you can memorize them before test day and save 10–15 seconds per question that you would otherwise spend re-reading them.

What DS actually tests: The temptation in DS is to solve the problem. Resist it. You never need a numerical answer — only a determination of whether the data is sufficient. Many DS traps are built around this distinction. A statement that gives you a value for x is sufficient. A statement that gives you two possible values for x may or may not be sufficient, depending on the question.

The most common DS mistake: Assuming that because a statement seems helpful, it is sufficient. Always ask: does this statement narrow the answer to exactly one possibility, or does it leave room for more than one?

Preparation approach: DS rewards systematic thinking over speed. Practice the "yes/no test" for yes/no questions: if Statement 1 always gives you "yes" or always gives you "no," it is sufficient — even if the answer is always "no." If Statement 1 sometimes gives "yes" and sometimes "no," it is insufficient. Drill this distinction until it is automatic.

---

2. Multi-Source Reasoning (MSR) — ~25% | ~5 Questions

Multi-Source Reasoning presents information across two or three tabbed sources — which may include text passages, tables, charts, or a combination — and asks you to synthesize, compare, or draw inferences from that information. A set of 2–4 questions is typically associated with a single set of sources.

The format: You see a prompt panel with multiple tabs. Each tab contains a different piece of information about the same scenario. Questions may ask you to identify what the sources collectively imply, what is supported by one source but contradicted by another, or what can be inferred when information from multiple tabs is combined.

What MSR actually tests: Information management and careful reading under time pressure. The sources are designed to be slightly complex — they contain both relevant and irrelevant information, and some questions are specifically designed to exploit the difference. The cognitive demand is managing multiple information streams simultaneously and knowing which tab to reference for each question.

Answer format variation: MSR includes both traditional multiple-choice questions and "either/or" format questions where you evaluate whether each of several statements is inferable or not inferable from the sources. Each statement is evaluated independently.

The most common MSR mistake: Spending too long reading all three tabs in full before answering any questions. This is time inefficient and often counterproductive — you will not retain everything, and you will be forced to re-read anyway. A better strategy: skim the tabs for structure and key data on first pass, then read carefully only the relevant portions when each specific question directs you.

Preparation approach: Practice with dense, multi-source reading materials before you encounter official MSR questions. Lawyer briefs, consulting case documents, and multi-section news articles all train the same skill: extracting the right piece of information from a complex, multi-part source efficiently.

---

3. Table Analysis (TA) — ~15% | ~3 Questions

Table Analysis presents a sortable data table — typically with several columns of numerical or categorical data — and asks you to evaluate a series of statements about that data. The table can be sorted by any column, which is a feature you are expected to use strategically.

The format: You see a data table with a sort function at the top of each column. Below the table are typically three to five statements in a "true/false" or "yes/no" format, each evaluated independently. You must determine whether each statement is supported or contradicted by the table data.

What TA actually tests: Numerical reasoning, data interpretation, and the ability to identify what a table does and does not show. Many statements in TA are carefully worded to distinguish between "the data supports this" and "this is plausible but not shown by the data." That distinction is critical.

Using the sort function strategically: The sort function is not decorative. Questions are often designed around sorted views of the data — for example, asking about the top 5 values in a column, or whether a trend holds across a certain range. Learn to sort the table immediately when a question references ranking, order, or relative magnitude.

The most common TA mistake: Eyeballing data rather than reading it precisely. "Approximately" and "exactly" are very different in TA. If a statement says "the value in column A is always greater than the value in column B," you need to verify every row, not scan for counterexamples. Missing one data point can flip your answer.

Preparation approach: Build comfort with data tables by regularly reading business reports, financial statements, and data-heavy news coverage. Practice extracting specific claims from tabular data quickly and accurately. The goal is precision, not speed — though precision practiced consistently will become fast.

---

4. Graphics Interpretation (GI) — ~15% | ~3 Questions

Graphics Interpretation presents a visual — a scatter plot, bar chart, pie chart, line graph, or other visual format — and asks you to complete one or two statements about it by selecting from drop-down answer menus.

The format: A graphic is displayed with a brief explanatory context. Below it are one or two sentence stems with blanks that you complete by choosing from a drop-down list of 3–5 options. Each blank is evaluated independently, and there is no partial credit — both blanks in a two-blank question must be correct to receive credit.

What GI actually tests: Visual data literacy and precise language. The sentences you complete are carefully worded, and the answer options often include choices that are close but not quite right — for example, the difference between "positively correlated" and "strongly positively correlated," or between "the highest" and "among the highest."

Common graphic types to master:

• Scatter plots: Questions typically involve correlation direction (positive/negative), outliers, or specific data point identification
• Bar charts: Questions often involve comparison, relative magnitude, or percentage calculations
• Pie charts: Questions usually involve proportion, fraction, or share of total
• Line graphs: Questions typically involve trends over time, rates of change, or intersection points

The most common GI mistake: Misreading the axis. GMAT graphics often use non-standard scales — logarithmic axes, dual-axis charts, or axes that do not start at zero. Always read both axes completely before interpreting any data point.

Preparation approach: Spend time reading charts in financial publications, annual reports, and data journalism. Practice articulating precisely what a chart shows and what it does not show — the language discipline this builds directly translates to GI performance.

---

5. Two-Part Analysis (TPA) — ~15% | ~3 Questions

Two-Part Analysis is the most flexible question type in Data Insights and can draw on quantitative reasoning, verbal reasoning, or both simultaneously. It presents a scenario and asks you to find two components of a solution — both of which must be correct to receive any credit.

The format: A question stem presents a problem or scenario, followed by a table with several rows of answer options. Two columns are labeled for Part 1 and Part 2 of your answer. You select one answer per column, and those two selections together constitute your complete response.

What TPA actually tests: Logical reasoning and the ability to satisfy multiple constraints simultaneously. The challenge is that Part 1 and Part 2 are not independent — they often interact. Selecting the right answer for Part 1 constrains what can be correct for Part 2. Many TPA questions are essentially constraint-satisfaction problems.

Quantitative vs. verbal TPA: Quant-based TPA questions typically involve equations with two unknowns, optimization problems, or scenarios where two related quantities must be determined. Verbal-based TPA questions often involve arguments where you must identify, for example, both a position a party holds and an assumption that position depends on.

The most common TPA mistake: Treating Part 1 and Part 2 as independent questions. They are not. Always check that your two selections are mutually consistent and collectively satisfy the full requirements of the question stem.

Preparation approach: TPA rewards systematic elimination. Before selecting answers, define clearly what constraints each part must satisfy, then work through the answer options methodically. Speed comes with practice — on your first pass, prioritize getting the logic right.

---

Section 3: Time Management — The 45-Minute Challenge

Twenty questions in 45 minutes works out to 2 minutes and 15 seconds per question. In practice, this average conceals enormous variation — some question types consistently take longer than others, and within each type, individual questions vary dramatically in complexity.

Realistic time benchmarks by question type:

Question Type	Target Time
Data Sufficiency	1:45 – 2:15 per question
Multi-Source Reasoning	4:00 – 6:00 per MSR set (2–4 questions)
Table Analysis	1:30 – 2:30 per statement set
Graphics Interpretation	1:30 – 2:00 per question
Two-Part Analysis	2:00 – 3:00 per question

The critical insight from these benchmarks: MSR sets are time-expensive upfront because you must read and process the source tabs before answering. But once you have internalized the sources, the individual questions within the set should move quickly. Do not treat each MSR question as a standalone — the reading investment pays dividends across the whole set.

The Triage Mindset

Not all 20 questions are equal in difficulty, and not all are equal in the time they demand. Experienced test-takers develop a triage instinct: they recognize within the first 30 seconds of a question whether it is going to be quick, moderate, or a time sink, and they budget accordingly.

If you are 60 seconds into a question and you have no clear direction, make your best educated guess and move on. One difficult question abandoned is not a problem. Three difficult questions that each consume 4 minutes while you scramble will cost you 6 minutes of time that you needed for questions you could have answered correctly.

Pacing Checkpoints

Build two checkpoints into your Data Insights section:

• After question 7: You should have approximately 32 minutes remaining
• After question 14: You should have approximately 16 minutes remaining

If you are behind either checkpoint, accelerate on the next block. If you are ahead, do not rush — bank the time for complex MSR sets later in the section.

---

Section 4: How Data Insights Fits Into Your Overall GMAT Score

This cannot be overstated: Data Insights is not a bonus section or a supplementary measure. Under the GMAT Focus Edition scoring model, your total score is a composite of three equally weighted section scores — Quantitative Reasoning, Verbal Reasoning, and Data Insights. Each section score ranges from 60 to 90 in 1-point increments.

What equal weighting means in practice:

A candidate scoring 85 in Quant, 83 in Verbal, and 72 in Data Insights will score meaningfully lower than a candidate scoring 80 in each section — even though the first candidate has two higher section scores. Imbalance hurts. The GMAT Focus scoring algorithm rewards consistency across all three sections.

The under-preparation penalty: Because Data Insights is newer and less represented in legacy prep materials, it is the section where the average score-to-preparation ratio is most distorted. Candidates who invest equal preparation across all three sections consistently outperform those who treat Data Insights as secondary — often by 20–30 total score points.

---

Section 5: A Targeted Data Insights Preparation Framework

Given its weighting and structure, Data Insights deserves a dedicated preparation lane — not leftover time after Quant and Verbal.

Phase 1: Orientation (Week 1–2)

• Complete one full Data Insights section from an official practice test with no time pressure. The goal is exposure, not performance.
• Work through all five question types with explanations, understanding the format of each before attempting any under timed conditions.
• Identify which two question types feel least intuitive. These become your primary focus areas.

Phase 2: Type-by-Type Drilling (Weeks 2–5)

• Dedicate one week to each of your two weakest question types. Work 8–10 questions per day in that type, reviewing every explanation regardless of whether you got the question right.
• Rotate through all five types on a weekly basis to maintain familiarity across the full section.
• Begin tracking your accuracy by question type. Your error log should show you whether your weak areas are improving.

Phase 3: Timed Integration (Weeks 5–8)

• Begin taking full 20-question, 45-minute Data Insights sections under real timed conditions.
• Track not just accuracy but time per question type. Identify where your time overruns are occurring.
• Practice the two checkpoint system: pause at questions 7 and 14 to assess your pacing.

Phase 4: Full Mock Integration (Ongoing from Week 6)

• Take complete GMAT Focus practice tests that include all three sections. Evaluate your Data Insights performance in the context of a full exam, when you are competing with mental fatigue from the other sections.
• Review your Data Insights errors within 24 hours of each mock test. Add new error patterns to your log.

---

Section 6: Common Data Insights Mistakes and How to Fix Them

Not using the sort function in Table Analysis. Many candidates treat the table as static. The sort function exists specifically to help you answer questions — use it every time a question involves ranking, ordering, or comparisons across rows.

Reading all MSR tabs exhaustively before answering. This wastes 90 seconds you cannot spare. Skim for structure first. Answer what you can. Return to specific tabs for specific questions.

Forgetting that DS answer choices are fixed. Memorize the five DS answer choices before test day. Not having to re-read them on every question saves 60–90 seconds across a section.

Treating Two-Part Analysis columns independently. Both selections must work together. Always validate that your two answers satisfy the complete requirements of the question before moving on.

Misreading chart axes in Graphics Interpretation. Spend 10 seconds reading both axes fully before interpreting any data. This single habit prevents the most common GI errors.

Under-preparing relative to Quant and Verbal. The most expensive mistake. Data Insights is one-third of your score. Treat it that way from Day 1 of your preparation.

---

Conclusion: Make Data Insights Your Competitive Advantage

Here is the counterintuitive opportunity in Data Insights: because most candidates under-prepare for it, genuine preparation creates an outsized advantage. In a section where the average candidate is walking in with 60% readiness, arriving at 90% readiness does not just improve your Data Insights score — it separates you from a significant portion of the competitive applicant pool.

The five question types are learnable. The time management is manageable. The formats, once practiced, become familiar. None of this requires exceptional mathematical ability or verbal brilliance — it requires deliberate, systematic preparation starting early enough to matter.

Data Insights is not the section to figure out on test day. It is the section to master in preparation, so that on test day, it is the 45 minutes where you are most confident, most efficient, and most in control.

Start there.

---

This guide reflects the GMAT Focus Edition format as of 2026. Question type weightings are approximate estimates based on reported test structures and may vary across individual exam administrations.

---