The Data Insights section has more variety than any other part of the GMAT®. Twenty questions, five different formats, forty-five minutes, and an on-screen calculator that sometimes helps and sometimes just gives you a false sense of confidence.
If you've ever stared at a Multi-Source Reasoning tab set wondering which of three tabs holds the detail that eliminates answer choice B, you know the feeling. DI questions can eat time in ways that Quant and Verbal don't, because the format itself adds complexity on top of the reasoning. A Quant question tests your math. A Verbal question tests your reading. A DI question tests your ability to navigate a format, find the right data, and then do the reasoning on top of it.
Guessing well on Data Insights is a specific skill. It's not the same as guessing on Quant, where you can often estimate or backsolve. It's not the same as guessing on Verbal, where trap answer patterns are fairly consistent across question types. DI has five question types, and each one has its own elimination logic. We'll walk through each of them.
The Ground Rule: No Partial Credit
DI questions with multiple parts, like Two-Part Analysis and Table Analysis, require you to get every component right. There's no partial credit. If a Two-Part Analysis asks you to select two values and you get one right and one wrong, that's a wrong answer.
This matters for guessing strategy. If you can nail one part of a two-part question but you're stuck on the other, don't spend three minutes hunting for the second piece. Make your best guess on the uncertain part, commit, and move on. The probability of getting full credit by guessing one part is low, but the time you save can go toward questions you can actually solve.
The same logic applies to Table Analysis, which presents a set of yes/no or true/false statements. You typically need all of them correct for credit. If you can confirm two of three statements but the third is opaque, guess it and move on rather than burning four minutes trying to verify something you're unsure about.
Data Sufficiency: The AD-BCE Framework
Data Sufficiency lives in Data Insights now, not Quant. The format gives you a question followed by two statements, and you have to determine whether the statements provide enough information to answer the question.
The standard framework for DS is AD-BCE. If Statement (1) is sufficient on its own, the answer is either A (only Statement 1) or D (both statements are sufficient independently). If Statement (1) is not sufficient on its own, the answer is either B, C, or E. This framework is valuable for guessing because it cuts your choices in half before you even evaluate Statement (2). If you can determine that Statement (1) is sufficient but you're not sure about Statement (2), you're choosing between A and D, a 50/50 instead of a 1-in-5.
If Statement (1) is clearly not sufficient, you're choosing among B, C, and E. In that scenario, C is often a trap answer. C says "both statements together are sufficient, but neither alone is." Test writers love C because it feels safe, like you needed all available information. But C is often wrong when one statement alone is actually enough.
If you're guessing between B, C, and E and you have no read on Statement (2), lean away from C. Pick B if Statement (2) seems substantial. Pick E if both statements seem thin or redundant.
The C Trap shows up most when both statements introduce numbers that look complementary. The test is betting you'll assume you need both. If you can evaluate either statement alone, do it. If you can't, guess around C.
One more DS guessing tip: on yes/no questions, if both statements give you the same answer (both say yes or both say no), that's usually not sufficient. The question is asking whether you can determine the answer, not what the answer is. A statement that says "x is positive" doesn't answer "Is x greater than 5?" even though it feels like it's pointing in a direction.
Multi-Source Reasoning: Tab Navigation as Elimination
Multi-Source Reasoning gives you two or three tabs of information, usually a mix of text, tables, and email exchanges. The questions ask you to evaluate statements based on the tabs.
The biggest time trap in MSR is reading all three tabs thoroughly before looking at the question. Don't do that. Read the question first, then scan the tabs for the specific information the question asks about. This sounds obvious, but under time pressure, a lot of students default to reading everything because it feels thorough. It's not thorough. It's expensive.
For guessing, MSR rewards elimination based on tab sourcing. If a statement references data that only appears in Tab 2, and you've scanned Tab 2 and can't find anything supporting that statement, eliminate it. You don't need to re-read every tab to rule out a statement that claims something the source material doesn't address.
MSR statements tend to fall into three categories. DIRECTLY SUPPORTED means the tab says exactly this, so keep it. DIRECTLY CONTRADICTED means the tab says the opposite, so eliminate it. NOT ADDRESSED means the tabs don't mention this at all, so eliminate it. The third category is where students lose time, because they keep searching for confirmation that a statement is wrong when the tabs simply don't cover the topic. If you've done a targeted scan and can't find the topic, treat it as not addressed and move on.
When you're stuck between two MSR statements and time is running out, look for the one that uses language from the tabs. Statements that paraphrase the source material are more likely to be supported than statements that introduce new framing or interpretation.
Two-Part Analysis: Pick the Easier Half First
Two-Part Analysis questions give you two related sub-questions, each with the same set of answer choices (or two different sets). You select one answer for each part.
The guessing strategy here is straightforward: solve the part you can, then use that information to narrow the other part. Many TPA questions are designed so that one part is easier than the other. If Part 1 asks for a value and Part 2 asks for a relationship, and you can calculate the value, do that first. Then check whether the value you found constrains Part 2.
If you can't solve either part, use the answer choices to eliminate. In a TPA with five answer choices per part, if you can rule out two choices for Part 1, you're down to three options. That's already better than random.
For TPA questions that involve calculations, the calculator can help with elimination. If an answer choice implies a value that doesn't match your rough calculation, cut it. You don't need precision to eliminate, just a sanity check.
When you're truly stuck with no elimination on either part, pick the same answer for both. It's not a strong strategy, but on a two-part question where both parts share the same answer choices, getting both right by picking the same value is more likely than getting both right by picking two different values at random. This is a last resort, not a plan.
Graphics Interpretation: Read the Axis First
Graphics Interpretation questions show you a chart, graph, or plot, then ask you to complete one or two statements using drop-down menus. The statements have blanks, and each blank has 3-5 options.
The guessing strategy for GI starts before you look at the question: read the axes and the legend first. A lot of GI elimination comes from understanding what the graph shows and what it doesn't. If a statement's blank offers a value that would require data not shown on the graph, eliminate it. For example, if the graph shows revenue by quarter and a blank offers values in terms of profit, you can't derive profit from revenue alone. Eliminate any option that requires information the graph doesn't provide.
The calculator is useful for GI, but use it for elimination, not precision. If a blank asks for a percentage change and the options are 5%, 15%, 25%, and 40%, you don't need to calculate the exact value. You need to estimate whether it's closer to 15 or 25. Rounding the values from the graph and doing rough division with the calculator is usually enough to eliminate two or three options.
Watch for units. GI questions love to switch units between the graph and the answer choices. If the graph shows thousands of dollars and the options are in dollars, or the graph uses months and the options use years, that's a trap. Convert mentally before calculating, and eliminate any option that doesn't match the question's units.
Table Analysis: Sort Before You Eliminate
Table Analysis gives you a sortable spreadsheet and a set of statements to evaluate as true or false (or yes/no). You sort the table by different columns to check each statement.
The guessing strategy for TA is about efficient sorting. Don't try to verify every statement by reading the entire table. Sort by the column the statement references, then check the top and bottom rows. For example, if a statement says "the company with the highest revenue also had the lowest profit margin," sort by revenue descending and check the top row's profit margin. Then sort by profit margin ascending and check the top row's revenue. If those two rows are the same company, the statement is probably true. If they're different companies, it's probably false.
When you're stuck on a TA statement and time is short, look at the extremes. Most TA statements make claims about maximums, minimums, or relationships between columns, and a quick sort by each relevant column usually gives you enough information to make an educated guess without reading every row.
If a statement makes a quantitative claim, like "more than half the companies had revenue above $500,000," sort by revenue and count the rows above the threshold. The calculator can help here if the table is large.
When to Guess on DI: The Bookmark System
Every section on the GMAT® gives you three bookmarks. You can bookmark a question you're unsure about, then return to it at the end of the section if you have time remaining.
On DI, use bookmarks strategically. The section has 20 questions in 45 minutes, which gives you about 2 minutes and 15 seconds per question. Some question types are naturally faster. Data Sufficiency questions can often be evaluated in under 90 seconds if you know the framework. Multi-Source Reasoning questions can eat 3-4 minutes if you get lost in the tabs.
If you hit the 2-minute mark on a DI question and you're not close to an answer, bookmark it and guess. Don't let one MSR question eat the time you need for two Graphics Interpretation questions that you could solve.
The review screen at the end of each section lets you revisit bookmarked questions and change up to three answers. If you finish with time remaining, go back to your bookmarked questions. But don't plan on having lots of review time. Most students tend to use every minute on the first pass.
Calculator Strategy: Elimination Tool, Not Confirmation Tool
The on-screen calculator in DI is basic. It can add, subtract, multiply, divide, and handle square roots. It can't graph, can't do statistical functions, and can't store variables.
Use the calculator for elimination, not for precision. If you're trying to decide between two answer choices and the difference is 2%, you probably need the calculator. If the difference is 20%, you can estimate. The most common calculator mistake on DI is using it to confirm an answer you've already determined through reasoning. If you've eliminated four of five choices through logic, don't calculate to confirm the fifth. Pick it and move on.
The second most common mistake is using the calculator when the question doesn't require it. Many DI questions test reasoning, not computation. If a question asks whether a statement is supported by the data, the answer usually comes from reading the data correctly, not from calculating.
Putting It Together
Guessing on Data Insights is about format awareness. Each question type has its own elimination logic, and the faster you can identify which logic to apply, the more time you save. For Data Sufficiency, use AD-BCE to cut your choices in half and watch for the C Trap. For Multi-Source Reasoning, read the question first and eliminate anything the tabs don't address. For Two-Part Analysis, solve the easier half first and use it to constrain the harder half. For Graphics Interpretation, read the axes and legend first, and estimate rather than calculate when the answer choices are spread out. For Table Analysis, sort by the relevant column and check the extremes.
Bookmark questions you're stuck on, guess before the 2-minute mark, and use the calculator for elimination rather than confirmation.
Guessing isn't a failure on Data Insights. It's a time management skill. The section has too much variety to solve everything, and the students who tend to score well are usually the ones who know when to let go.
Frequently Asked Questions
How many DI questions can I guess on?
You can guess on as many as you need to. The adaptive scoring algorithm generally weighs question difficulty, so missing questions you'd never have solved anyway probably won't hurt your score much. What can hurt more is spending four minutes on a question you can't solve and then running out of time for questions you could have solved.
Should I guess the same letter on every DI question?
DI questions don't all use lettered answer choices. Data Sufficiency has the standard A through E options, but other question types use drop-down menus, yes/no selections, or two-part selectors. Letter-based guessing strategies don't apply across most of the section.
Is the calculator available for all DI questions?
Yes. The on-screen calculator is available throughout the entire Data Insights section. It's not available during Quant or Verbal.
What's the hardest DI question type to guess on?
Multi-Source Reasoning is usually the hardest to guess on because the questions are specific to the tab content. There's no general elimination framework that works across all MSR questions. You have to engage with the source material, even for a guess.
Should I save DI for last when choosing my section order?
Some students prefer to take DI when they're freshest, since it requires switching between formats. Others prefer to take it last so they can spend remaining energy on the section with the most variety. There's no universally correct answer. Try different orders on practice tests and see what works for you.
Want to learn even more?
Guessing on Data Insights is one part of a broader timing and pacing strategy. If you haven't read our guide to when to let go of a GMAT® question, that's the companion piece to this post.
For the full pacing framework across all three sections, our GMAT® timing strategy guide walks through checkpoints, speed benchmarks, and common pacing traps.
If you want the overview of guessing across all sections, our GMAT® guessing strategies guide covers the general elimination framework and section-specific tactics.
For the Data Insights section specifically, our GMAT® Data Insights complete guide covers each question type in more depth, with examples and practice approaches.
And if you want to hear Isaac talk through these strategies in his own words, the podcast is available on Spotify, Apple Podcasts, and YouTube. Several episodes cover DI strategy, calculator usage, and test-day decision-making.