GMAT Data Insights: Complete Section Guide (Updated 2026)
Quick Takeaways
- Mindset: DI is a data literacy test, not a calculation test — the calculator is a trap as much as a tool.
- DS Strategy: Determine if you can solve it, not what the answer is.
- MSR Strategy: Skim all tabs first, build a mental map, then read the questions.
- TA Strategy: Sort the column before doing anything else.
- Traps: Confusing correlation with causation and treating 'Cannot Determine' as 'False' are the #1 score killers.
- Pace: ~1:30 for DS, ~2:00 for GI and TA, ~3:30–4:00 per MSR prompt.
- Scoring: DI = one-third of your total GMAT score. Don't neglect it for Quant.
What is Data Insights?
The GMAT Data Insights (DI) section is a 45-minute, 20-question section that measures your ability to analyze, interpret, and synthesize data from multiple sources to make sound decisions. It's designed to simulate real-world business scenarios where information arrives in multiple formats — tables, charts, text passages, emails, and spreadsheets — and where decisions must be made quickly and accurately.
Unlike the Quantitative section, DI is not primarily about computation. The on-screen calculator is available for every question, but reaching for it on every problem is a trap — the section rewards structured thinking, pattern recognition, and logical judgment far more than arithmetic ability. The student who estimates smartly and thinks clearly will almost always outscore the one who calculates precisely but slowly.
DI is one of three sections in the GMAT Focus Edition alongside Quantitative Reasoning and Verbal Reasoning. Each section contributes equally — one-third — to your total score on the 205–805 scale, making DI just as important as Quant or Verbal for your final result.
| Feature | Details |
|---|---|
| Number of Questions | 20 |
| Time Limit | 45 minutes |
| Question Types | DS, MSR, Table Analysis, Graphics Interpretation, Two-Part Analysis |
| Scoring Range | 60–90 |
| Adaptive? | Yes — difficulty adjusts after every question |
| Calculator? | Yes — available for all questions |
| Avg. Time Per Question | ~2 min 15 sec (varies significantly by type) |
What Changed from the Old GMAT?
If you've studied for the pre-Focus GMAT, Data Insights will feel familiar in parts and brand new in others. The old exam split data skills across two separate areas: Integrated Reasoning (IR), a 12-question standalone section scored on a 1–8 scale that most schools barely considered, and Data Sufficiency, embedded inside the 37-question Quantitative section.
The GMAT Focus Edition merged and upgraded both into a single, cohesive DI section — and critically, gave it equal weight in the total score. This is the most important structural change in the new exam: skills that were previously tested in a low-stakes appendix now count for a full third of your score.
| Feature | Old GMAT | GMAT Focus Edition |
|---|---|---|
| Data Sufficiency | Part of Quant (37 questions) | Standalone in DI section |
| Integrated Reasoning | Separate section, 12 Qs, 1–8 scale | Merged into DI |
| Score Impact | IR not included in total score | DI = 1/3 of total score |
| Calculator | IR only | All DI questions |
| Question Types | DS + 4 IR types | 5 unified DI types |
The 5 DI Question Types
DI contains five distinct question types, each testing a different facet of data literacy. Understanding what each type demands — and what traps it sets — is the foundation of a strong DI score.
| Question Type | Approx. Questions | Format |
|---|---|---|
| Data Sufficiency | 5–8 | Standalone, 5 fixed answer choices |
| Graphics Interpretation | 4–6 | 1 graphic, 2 fill-in-the-blank dropdowns |
| Table Analysis | 4–6 | 1 sortable table, multiple True/False items |
| Multi-Source Reasoning | 5–6 | 2–3 tab prompts, 3 questions per prompt |
| Two-Part Analysis | 2–4 | 1 problem, 2 linked answers from same list |
1. Data Sufficiency (DS)
Data Sufficiency is the most common question type in DI and the one with the steepest learning curve for first-time test-takers. You're given a question stem and two statements of data. Your task is not to solve the problem — it's to determine whether the given information is sufficient to solve it. This is a fundamentally different skill from problem-solving. Most students instinctively try to find the answer; DS rewards students who can identify whether a unique answer is findable at all.
DS Answer Choices — Memorize These (A) Statement (1) alone is sufficient, but (2) alone is not (B) Statement (2) alone is sufficient, but (1) alone is not (C) Both statements together are sufficient, but neither alone is (D) Each statement alone is sufficient (E) Neither statement alone nor together is sufficient
DS Strategy Use the AD/BCE method: Evaluate Statement (1) alone first. Sufficient → answer is A or D. Not sufficient → answer is B, C, or E. This cuts your decision tree in half immediately. Never assume values: A statement is sufficient only if it yields a unique answer for all valid inputs — not just the one you tested. Yes/No vs. What is: For Yes/No questions, 'always yes' or 'always no' counts as sufficient. 'Sometimes yes, sometimes no' does not. Watch integer constraints: A statement that seems insufficient for real numbers may become sufficient when the variable is restricted to integers. Stop at sufficiency: Once you confirm a unique answer exists, move on. Fully solving wastes 30–60 seconds per question across the section.
Worked Example: DS Is x > 0? (1) x² = 9 (2) x³ > 0 Deconstruct: We need to determine the sign of x — not its value. Statement (1) alone: x² = 9 means x = 3 or x = −3. Both are possible, so we cannot determine the sign. Insufficient → eliminate A and D. Statement (2) alone: x³ > 0 means x must be positive (a negative number cubed is always negative). So x > 0 definitively. Sufficient → answer is B. The Trap: Statement (1) is designed to make students write x = 3 and stop — forgetting the negative root. DS is built to punish exactly this kind of assumption. Final Answer: B
2. Multi-Source Reasoning (MSR)
MSR presents information across two or three clickable tabs that can contain text passages, tables, charts, emails, or any combination. Each MSR prompt comes with three questions that require you to synthesize information from across the tabs — and each of those three questions is scored independently. MSR is the most reading-intensive question type in DI and where most students lose the most time.
MSR Strategy Skim all tabs first (~60 sec): Before reading any question, build a mental map of what lives where. You need to know where to look — not memorize every detail. Each question is scored independently: Getting Q1 wrong doesn't hurt Q2 or Q3. Don't waste time revisiting a question you've already moved past. 'Cannot Determine' ≠ 'False': If a source doesn't address a statement, the answer is 'Cannot Determine' — not False. False means the source contradicts it. This is the single most common MSR error. Inference standard is high: Only mark 'True' or 'Yes' if the information is 100% logically supported. Never infer beyond what the data explicitly says. Budget ~3:30–4:00 per prompt: Not per question — per the entire three-question set. Front-load time on reading the tabs.
3. Table Analysis
Table Analysis gives you a sortable spreadsheet-style table followed by a series of True/False or Yes/No statements evaluated using only the data in the table. The sortable column feature is your most powerful tool — and most students significantly underuse it.
Table Analysis Strategy Sort aggressively: Don't manually scan the table. Sort the relevant column to find medians, outliers, and maximums in seconds. True/False logic is strict: If the data doesn't fully and unambiguously support a statement, it's False — not 'probably true'. Watch the % vs. absolute trap: The highest percentage share row is not always the row with the highest absolute value. This appears in almost every TA question. Check column headers carefully: The most common TA errors come from misreading what a column actually represents — units, time periods, or whether values are cumulative vs. per-period. Year-over-year ≠ cumulative: Always confirm whether the table shows period values or running totals before answering any trend-based statement.
4. Graphics Interpretation
Graphics Interpretation presents a single visual — a bar chart, line graph, scatter plot, pie chart, or box-and-whisker plot — followed by two statements with dropdown blanks to fill in. The entire data universe is the graphic itself. GI is often the fastest question type when you read the chart correctly upfront, and the easiest to get wrong when you don't.
| Chart Type | What It Tests | Key Watch-Out |
|---|---|---|
| Bar Chart | Category comparison | Stacked vs. grouped bars |
| Line Graph | Trends over time | Dual axes with different units |
| Scatter Plot | Correlation direction and strength | Outliers vs. trend line |
| Pie Chart | Part-to-whole proportions | Absolute vs. percentage values |
| Box Plot | Distribution and skewness | Median position within box = skew direction |
GI Strategy Read title and axes before anything else: Most errors come from misidentifying what is being measured, not from misreading the data. Check for broken axes: A y-axis that doesn't start at zero makes small differences appear dramatically large. This is a deliberate GMAT trap. Dual axes require extra caution: When two y-axes with different units appear on the same chart, confirm which data series belongs to which axis. For scatter plots: Identify the trend line direction first, then note which labeled points are outliers relative to it. For box plots: The median line's position within the box tells you skewness — median closer to Q1 means right-skewed; closer to Q3 means left-skewed.
5. Two-Part Analysis
Two-Part Analysis presents a single problem requiring two related answers, both selected from the same list of options. Correct answers can and do appear in the same row. TPA questions can be quantitative (systems of equations, optimization, trade-off) or verbal/logical (constraint satisfaction, argument evaluation). Both parts must be correct to receive any credit — there is no partial credit.
Two-Part Strategy Read both columns before solving: Understanding how the two parts relate is more important than solving either one individually. Test answer choices systematically: Plug choices back in rather than deriving from scratch — elimination is faster than construction for most TPA questions. Correct answers can and do appear in the same row: Don't eliminate a row because you already selected it for the first part. For verbal TPA: Identify the core constraint or condition linking both parts before evaluating any options. Verify both parts satisfy ALL conditions: The most common TPA error is confirming one part without checking it doesn't violate the other part's condition.
Skills Tested in DI
DI is not primarily a math test. It rewards a specific combination of skills that business schools use to predict success in case-based learning environments.
- Data Interpretation: Reading what charts, tables, and graphs are actually showing — including units, scales, and time periods. Not what you think they show, but what they actually show.
- Data Synthesis: Combining information from multiple sources to reach a single, well-supported conclusion. This is the core skill of MSR and what separates 75s from 85s.
- Logical Reasoning: Drawing only conclusions strictly supported by evidence. Knowing when data is ambiguous is just as important as knowing when it's definitive.
- Quantitative Reasoning: Applying arithmetic, percentages, ratios, and basic statistics — conceptually more than computationally. Heavy calculation is rarely required.
- Decision Making Under Ambiguity: Recognizing when you have enough information to act and when you don't. The core skill of DS, and a foundational business skill.
How DI is Scored
DI is scored on a 60–90 scale in one-point increments and contributes equally to your total GMAT score alongside Quant and Verbal. The section uses a question-level adaptive algorithm — each question's difficulty adjusts based on your performance on previous questions. Getting harder questions right is how you move from the 70s into the 80s.
Partial credit mechanics vary by question type. MSR questions are scored individually — each of the three questions in a prompt is independent, so getting one wrong doesn't affect the others. Two-Part Analysis requires both parts to be correct for any credit. Table Analysis and Graphics Interpretation items are evaluated per statement.
| Scoring Fact | Detail |
|---|---|
| Score Range | 60–90 |
| Contribution to Total | ~33% (equal weight with Quant and Verbal) |
| Adaptive Level | Question-by-question |
| Guessing Penalty | None — always answer every question |
| TPA Partial Credit | No — both parts must be correct |
| MSR Partial Credit | Yes — each question scored independently |
| A Score of 80+ | Top ~25% globally |
Time Management Strategy
With 20 questions in 45 minutes, your average is 2 minutes 15 seconds per question. But treating every question equally is how you run out of time. Different question types have fundamentally different time demands — the key is building a per-type budget and sticking to it.
Accuracy matters throughout the section — there is no 'safe' section where you can rush to save time for later. Build a consistent per-type pace rather than front-loading caution on the early questions.
| Question Type | Target Time | Key Note |
|---|---|---|
| Data Sufficiency | ~1:30 per question | Stop the moment sufficiency is clear — don't solve |
| Graphics Interpretation | ~2:00 per question | Invest time in chart-reading, not answer-deliberation |
| Table Analysis | ~2:00 per question | The sort function saves 30–60 sec per question |
| Multi-Source Reasoning | ~3:30–4:00 per prompt | Budget for the full 3-question set, not each question |
| Two-Part Analysis | ~2:30–3:00 per question | Verification step is essential and takes time |
Pacing Checkpoints After Q7: ~30 minutes remaining. If behind, you're over-spending on DS or GI. After Q13: ~16 minutes remaining. If behind, MSR has likely consumed your buffer. After Q18: ~5 minutes remaining. Use this to resolve any flagged questions. Flag and move on: After 3 minutes on any question, make your best guess and move. A forced guess plus 3 minutes saved beats a right answer costing 5 minutes. Never leave a question blank: There's no penalty for wrong answers — always select something before time runs out.
A Special Note for Indian Test-Takers
For Indian MBA aspirants — particularly those targeting ISB, INSEAD, LBS, or NUS — DI often presents a specific challenge: strong quantitative instincts combined with a tendency to over-calculate and under-read. The on-screen calculator feels like a safe anchor. On DI, it's a time trap.
Rebalance Your Approach Resist the calculator: Practice estimation until it feels as natural as calculation. Most DI questions are designed to be solved without it. Build MSR reading stamina: Reading dense, multi-tab passages under time pressure is one of the more common challenges. Read 20–30 minutes daily from The Economist or Harvard Business Review — not for content, but for reading-under-pressure conditioning. Don't let DS become algebra practice: DS questions trigger full equation-solving instincts. Train yourself to stop the moment sufficiency is established — not when the value is found. Your strength is structure: Indian test-takers typically excel at pattern recognition and systematic elimination. Lean into this for TPA and TA — these are the most process-driven question types in DI.
Universal Strategies for DI Success
These strategies apply across all five question types. Master these fundamentals before diving into type-specific drilling — they are the foundation everything else is built on.
Read the question before the data. For DS, knowing what you're solving for changes how you evaluate the statements. For GI and TA, knowing the question before studying the chart or table tells you exactly where to look — and what to ignore.
Correlation is not causation. The GMAT tests this distinction explicitly and repeatedly — in GI scatter plots, in MSR text passages, and in TPA verbal questions. Two variables moving together is never sufficient to conclude one causes the other.
Eliminate with a reason. When stuck between two options, find a specific logical reason one is definitively better. If you can't articulate it, you haven't found the right answer yet.
| Aspect | What to Do | What to Avoid |
|---|---|---|
| Calculator | Use for complex arithmetic only | Reaching for it on every question |
| MSR | Skim tabs first, then read questions | Reading questions before knowing the sources |
| DS | Stop at sufficiency, don't solve fully | Fully solving to verify the answer |
| TA | Sort the column before scanning | Manually reading rows top to bottom |
| GI | Read axes and title first | Interpreting the visual before understanding the units |
| TPA | Verify both parts satisfy all conditions | Confirming one part and assuming the other follows |
4-Week DI Study Plan
This plan assumes roughly 1–1.5 hours per day dedicated to DI alongside your Quant and Verbal prep.
| Week | Focus | Key Tasks |
|---|---|---|
| Week 1 | Learn the Formats Focus: Build foundational understanding of all 5 question types | Study DS answer choice framework and the AD/BCE elimination method Learn to read all five chart types cold: bar, line, scatter, pie, box plot Practice 10–15 untimed questions per type to build pattern recognition Take the OpenPrep DI Diagnostic to identify your baseline weaknesses by question type |
| Week 2 | Type-by-Type Timed Practice Focus: Build speed and accuracy per question type | 10 timed questions per type under realistic time pressure Drill DS: practice recognizing sufficiency vs. insufficiency without solving Drill MSR: practice the 60-second tab-skim before reading any question Log every error — categorize as 'content gap' or 'process error' |
| Week 3 | Full Section Simulation Focus: Simulate real test conditions across all 5 types together | Complete 2–3 full 20-question DI sections under timed conditions Analyze time logs: where are you consistently over-budget? Focus extra sessions on your two weakest question types from Week 2 Practice the skip-and-flag strategy — know when to cut your losses and move on |
| Week 4 | Polish and Full Mocks Focus: Refine weak spots and build exam-day confidence | Take 2 full GMAT mocks (all 3 sections) and review DI performance in detail Revisit error log from Weeks 1–3 — look for recurring patterns, not individual mistakes Practice estimation techniques to break calculator dependency Final timed DI section 2 days before exam — no new material, only consolidation |