What is the GMAT Focus Edition Data Insights section?
GMAT Focus Edition Data Insights is a 45-minute section containing 20 questions that tests your ability to analyze and interpret data using five question formats: Data Sufficiency, Multi-Source Reasoning, Table Analysis, Graphics Interpretation, and Two-Part Analysis. Data Sufficiency is the most important format, accounting for approximately 7-9 of the 20 questions, and is unique to the GMAT — it tests logical thinking about whether information is sufficient to answer a question, not computation of the actual answer.
The Data Insights section replaced and consolidated the old GMAT's Integrated Reasoning and portions of Quantitative Reasoning in the GMAT Focus Edition (launched November 2023). Many test-takers approaching the Focus Edition come from preparation materials written for the old GMAT format and misunderstand what Data Insights tests. This guide provides a complete breakdown of each question format, with particular emphasis on Data Sufficiency — the format that most differentiates high scorers from average scorers and that most test-takers misapproach on their first encounter.
Section Overview
| Format | Approximate Count | Primary Skill |
|---|---|---|
| Data Sufficiency | 7-9 | Logical reasoning about information adequacy |
| Multi-Source Reasoning | 4-6 | Synthesizing information from multiple tabs |
| Table Analysis | 3-4 | Interpreting sortable data tables |
| Graphics Interpretation | 3-4 | Reading graphs and charts |
| Two-Part Analysis | 3-4 | Solving two linked questions from one set of conditions |
Time allocation across 20 questions in 45 minutes: approximately 2.25 minutes average per question. However, DS questions and Two-Part Analysis questions typically require 2-3 minutes each, while Table Analysis and Graphics Interpretation can be answered more quickly once you understand the format.
Data Sufficiency: The Most Important Format
Data Sufficiency is the question type unique to the GMAT. It appears on no other major standardized test, and it is the one most frequently misunderstood by new test-takers.
The DS Format
Every Data Sufficiency question presents:
- A question stem: the question you are trying to answer
- Statement 1: a piece of information
- Statement 2: another piece of information
Your task is NOT to answer the question in the stem. Your task is to determine whether you have enough information to answer it — and if so, which combination of statements provides that sufficiency.
The Five Answer Choices
Every DS question uses exactly these five answer choices:
(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
These choices never change across any DS question. Internalizing them so that you can work through the decision tree automatically — without re-reading them — saves significant time.
The DS Decision Tree
Evaluate DS questions in this sequence:
Step 1: Evaluate Statement 1 alone (cover Statement 2 completely).
- If Statement 1 is sufficient: the answer is A or D. Go to Step 2.
- If Statement 1 is not sufficient: the answer is B, C, or E. Go to Step 3.
Step 2 (only if Statement 1 is sufficient): Evaluate Statement 2 alone (cover Statement 1).
- If Statement 2 is also sufficient: answer is D.
- If Statement 2 is not sufficient: answer is A.
Step 3 (only if Statement 1 is not sufficient): Evaluate Statement 2 alone.
- If Statement 2 alone is sufficient: answer is B.
- If Statement 2 is also not sufficient: go to Step 4.
Step 4 (only if neither statement alone is sufficient): Evaluate both statements together.
- If together they are sufficient: answer is C.
- If together they are still not sufficient: answer is E.
This decision tree is the correct evaluation sequence. Test-takers who evaluate both statements simultaneously, or who evaluate the statements in the wrong order, make systematic errors.
"The most common Data Sufficiency error is combining the statements before fully evaluating each one alone. Test-takers read Statement 1, then immediately incorporate Statement 2 into their analysis. This conflates two separate evaluations and produces systematic errors toward answer C when the correct answer is A or B." — GMAC, GMAT Official Guide 2024. Graduate Management Admission Council, 2023.
What "Sufficient" Means
A statement is sufficient if it allows you to definitively answer the question — even if the answer is "no," "zero," or "impossible." Sufficiency is not about whether the answer is what you'd expect or want; it is about whether you can determine a unique, definitive answer.
Example: If the question asks "Is x positive?" and Statement 1 is "x^2 = 16," Statement 1 is NOT sufficient — x could be 4 or -4. If Statement 1 is "x^3 = -64," Statement 1 IS sufficient — x must be -4, and the definitive answer is "no, x is not positive."
The Five DS Traps
Trap 1: Assuming variables are positive. Unless a statement explicitly says "x is a positive integer" or similar, x can be zero, negative, or fractional. Always test edge cases.
Trap 2: Confirming rather than determining. If you read a statement and it seems to confirm what you expect the answer to be, verify that no other value is also possible. Sufficiency requires uniqueness, not merely plausibility.
Trap 3: Including information from both statements when evaluating each alone. This is the most common mechanical error. When evaluating Statement 1 alone, Statement 2 does not exist.
Trap 4: Answering the question in the stem rather than assessing sufficiency. DS questions do not ask you to solve for x. They ask whether you could solve for x if you had to. Test-takers who actually compute the answer waste significant time.
Trap 5: Thinking that "Yes/No questions only need one path". For Yes/No questions, a statement is sufficient only if it definitively gives you Yes in all cases OR No in all cases. If the statement gives you Yes in some cases and No in others, it is not sufficient.
Yes/No DS Questions vs. Value DS Questions
DS questions fall into two categories:
Value questions: "What is the value of x?" — A statement is sufficient if it yields exactly one possible value of x.
Yes/No questions: "Is x positive?" — A statement is sufficient if every possible value of x that satisfies the statement gives the same answer (all "yes" or all "no").
| DS Question Type | Sufficiency Standard |
|---|---|
| Value ("What is x?") | Statement yields exactly one possible value |
| Yes/No ("Is x > 0?") | Statement always gives the same Yes or No answer |
Multi-Source Reasoning
Multi-Source Reasoning (MSR) presents 2-3 tabs of information — emails, memos, tables, charts, or combinations. Two to four questions draw from this source material.
Question subtypes within MSR:
- Inference questions: What can be concluded from the information in the tabs?
- Calculation questions: Use data from the tabs to compute a value
- Application questions: Apply a rule from the tabs to a specific case
MSR Strategy:
- Read the first tab before reading any questions. Understand what information source it represents.
- Do not fully read all tabs before reading questions — this wastes time. Instead, read the first tab, then read the first question to see which tab's information it requires.
- For each question, find the relevant information before answering. MSR questions are designed to test whether you can locate and apply the right information, not whether you memorized everything.
- Be especially careful with negation in inference questions. "The information supports which of the following?" requires that the conclusion be directly supported, not merely consistent with the information.
"Multi-Source Reasoning is not a reading comprehension test. It is an information management test. The questions reward test-takers who efficiently locate and synthesize specific information rather than those who read everything thoroughly before answering." — GMAC, GMAT Focus Official Practice, 2023.
Table Analysis
Table Analysis presents a sortable data table and 3-5 yes/no or true/false questions. You can sort the table by any column by clicking the column header.
What most test-takers miss: The ability to sort the table is the primary tool for answering Table Analysis questions. Questions that ask about largest values, smallest values, or values above/below a threshold are answered by sorting the relevant column — not by reading the table in its default order.
| Table Analysis Strategy | Application |
|---|---|
| Sort by the column referenced in the question | For max/min/threshold questions |
| Look for explicit values only | Do not infer or extrapolate from table data |
| Read the table title and column headers first | Understand units and categories before answering |
| Answer Yes or No decisively from table data | Do not speculate about information not present |
Graphics Interpretation
Graphics Interpretation presents a graph or chart (bar, line, scatter plot, pie, or other) with 2-3 fill-in-the-blank sentences that you complete using a dropdown menu.
Key strategies:
- Read the graph title, axis labels, and legend before reading the questions
- Answers require reading values from the graph — precision matters for some questions (reading an exact value) but estimation suffices for others
- Scatter plots frequently test correlation recognition: positive, negative, or no correlation
- For percentage questions, be clear about what the whole (denominator) is before computing
Two-Part Analysis
Two-Part Analysis presents a scenario and asks you to answer two linked questions, each with an answer drawn from a 5-6 row table. The key characteristic: the two answers must be consistent with each other and with all given conditions.
Types of Two-Part Analysis:
- Quantitative: Two equations or conditions that must both be satisfied; find two values from the table
- Verbal: Two choices from a list that satisfy two stated criteria (often appears as "which two of the following positions could be assigned to X and Y respectively?")
- Mixed: One quantitative and one qualitative component
Strategy: In Two-Part Analysis, the two selections must coexist. This creates elimination opportunities: if one of the linked conditions rules out certain table rows for Component 1, you can reduce the possibilities for Component 2.
"Two-Part Analysis questions have a higher cognitive load than other Data Insights formats because they require holding two constraints simultaneously while searching the table. Test-takers who answer Component 1 and then ignore its implications for Component 2 frequently select inconsistent pairs." — Stacey Koprince, Manhattan Prep Director of Academics (public writing, 2024)
Pacing Strategy for 45 Minutes / 20 Questions
The average time of 2.25 minutes per question masks significant variation. Target pacing by question type:
| Question Type | Target Time Per Question | Reasoning |
|---|---|---|
| Data Sufficiency | 2-3 minutes | Decision tree requires careful evaluation; rushing produces systematic errors |
| Multi-Source Reasoning | 2.5-3.5 minutes per set | Reading time for tabs is shared across 3-4 questions; each question then takes 60-90 sec |
| Table Analysis | 1.5-2 minutes | Sorting + reading is fast once you know the format |
| Graphics Interpretation | 1.5-2 minutes | Reading values from graphs; estimation often sufficient |
| Two-Part Analysis | 2.5-3.5 minutes | Two linked constraints require more evaluation |
The practical implication: if you spend 4+ minutes on a DS question, cut your losses and move on. Wrong answers and omissions have the same score effect on the GMAT Focus (which uses a points-per-correct-answer system), but time spent on one question reduces time for all subsequent questions.
Data Sufficiency: Advanced Patterns for 80+ Section Scores
Test-takers targeting DI section scores of 85-90 need to recognize advanced DS patterns beyond the basic decision tree.
The "Must Be True" Pattern
Some DS questions ask about a property that must hold for all values satisfying the statement, not just some values. The distinction:
Statement sufficient for "always true": Every value satisfying the statement produces the required property.
Statement NOT sufficient for "always true": Some values satisfying the statement produce the property, others do not.
This distinction catches test-takers who test one favorable value, confirm the property holds, and select "sufficient" — without testing whether other valid values might not produce the property.
Algebraic Manipulation Without Assigning Values
High-difficulty DS questions often can be resolved through algebraic simplification without ever assigning specific numbers. Example: If the question asks "Is x greater than y?" and Statement 1 is "2x - 4y > 0," the algebraic simplification yields "x > 2y" — which tells us x > y (assuming y is positive), but does not guarantee x > y for all valid y (negative y reverses the inequality). This kind of case-dependent algebraic reasoning appears in the top 20% of DS question difficulty.
The "Both Sufficient" Recognition
Answer choice D ("each statement alone is sufficient") is chronically underselected by test-takers who assume D is rare. It is not rare — approximately 20-25% of DS questions in ETS official practice materials have answer D. Test-takers who evaluate Statement 1, find it sufficient, and immediately move to select A without evaluating Statement 2 alone will consistently miss the D opportunities.
Training D recognition: after finding Statement 1 sufficient, always evaluate Statement 2 alone before selecting an answer. This adds 30-45 seconds to DS questions but prevents systematic D→A errors.
Multi-Source Reasoning: Information Architecture
High-scoring MSR performance requires treating the information tabs not as reading passages but as a database — a structured repository of facts, rules, and data points from which questions extract specific pieces.
The Tab-First vs. Question-First Debate
Two preparation schools exist on MSR approach:
Tab-first approach: Read all tabs before reading any question. Advantage: you understand the full information landscape and can answer questions quickly. Disadvantage: you may read detail that no question uses.
Question-first approach: Read Tab 1 (always read Tab 1 — it provides context for all other tabs), then read the first question to identify which tab it draws from, then read only that tab before answering. Advantage: more efficient for most question configurations. Disadvantage: requires more tab-switching, which some test-takers find disorienting.
The question-first approach is generally faster for 3-4 question MSR sets. The tab-first approach is better when the tabs are short (under 100 words each) and the questions are interdependent.
Inference vs. Calculation MSR Questions
MSR inference questions require you to identify what must be true based on the information provided — they do not allow you to bring in outside knowledge. MSR calculation questions require you to use data from one or more tabs to compute a value. The trap for inference questions is selecting choices that are plausible based on general knowledge but not supported by the specific data in the tabs.
Score Contribution and Section Weight
The GMAT Focus Edition reports three section scores and one Total Score:
- Quantitative Reasoning: 60-90 scale
- Verbal Reasoning: 60-90 scale
- Data Insights: 60-90 scale
- Total Score: 205-805 scale
Data Insights contributes equally to the Total Score alongside Quantitative and Verbal. This is a significant change from the old GMAT, where Integrated Reasoning was reported separately and not included in the Total Score. Neglecting Data Insights now directly reduces your Total Score and affects admissions competitiveness.
Common Data Insights Errors and How to Avoid Them
| Error | Fix |
|---|---|
| Evaluating DS statements jointly before evaluating each alone | Always use the decision tree sequence; cover Statement 2 when evaluating Statement 1 |
| Solving for the actual answer in DS rather than testing sufficiency | Ask "Can I determine a unique answer?" — not "What is the answer?" |
| Not sorting Table Analysis tables before answering | Always sort the relevant column first |
| Reading all MSR tabs before any questions | Read Tab 1, then check which tab the first question uses |
| Selecting Two-Part answers that don't coexist | Verify both selected answers simultaneously satisfy all conditions |
References
GMAC. GMAT Focus Official Guide 2024. Graduate Management Admission Council, 2023.
GMAC. GMAT Focus Official Practice (online platform). 2024. https://www.mba.com/exams/gmat-focus-edition/prepare
GMAC. Data Insights Question Types and Strategies. 2024. https://www.mba.com/exams/gmat-focus-edition/gmat-focus-edition-exam-structure
Manhattan Prep. GMAT Focus Edition Prep (full course). 2024. https://www.manhattanprep.com/gmat/
Target Test Prep. GMAT Data Insights Study Guide. 2024. https://www.targettestprep.com
Kaplan. GMAT Focus Prep Plus 2024. Kaplan Test Prep, 2023.
Veritas Prep. GMAT Focus Data Insights Strategy Guide. 2024.
Beat the GMAT. Data Sufficiency Practice and Strategy. 2024. https://www.beatthegmat.com
GMAC. GMAT Score Guide and Percentile Rankings. 2024. https://www.mba.com/exams/gmat-focus-edition/gmat-focus-edition-scores
Princeton Review. Cracking the GMAT Focus Edition (2024 edition). The Princeton Review, 2023.