4 conflict patterns that decide your GMAT Focus…

The GMAT Focus edition of the Data Insights section contains a question family built entirely around comparing multiple data sources. On the screen it looks deceptively simple: three small tabs sit above a short prompt, each tab opens a different exhibit, and the candidate is asked to draw a conclusion that no single tab supports on its own. Underneath that clean interface sits the most structurally demanding question type in the section, and the one that most reliably separates a 645 scorer from a 705 scorer. The family is called Multi-Source Reasoning, it appears in sets of two to three linked items, and it is the only GMAT Focus question type where reading order, tab discipline, and note-taking carry more weight than arithmetic speed.

Candidates who treat each tab as a self-contained mini-passage tend to answer one of the items correctly and the other two incorrectly, because the items are engineered to be cross-dependent. A fact that lives only on Tab 2 has to be combined with a column header on Tab 1 and a footnote on Tab 3 before the stem resolves. The skill the test is measuring is therefore not reading, not arithmetic, and not chart interpretation in isolation; it is the ability to hold three small exhibits in working memory long enough to merge them into a single mental model, then answer three questions against that merged model without re-reading the original exhibits from scratch. This article is the senior-tutor walkthrough of that exact skill: what the question family is testing, how to triage the tabs in the first 30 seconds, what to write down, and the four conflict patterns that decide the answer on the hardest items.

What the GMAT Focus actually measures with a Multi-Source Reasoning set

Multi-Source Reasoning is one of the four item families in the GMAT Focus Data Insights section, alongside Data Sufficiency, Table Analysis, Graphics Interpretation, and Two-Part Analysis. The section as a whole is scored on a scale from 60 to 90 and contributes alongside the Quant and Verbal scaled scores to the composite GMAT Focus total. Inside the section, Multi-Source Reasoning is the only family that ships more than one question per stimulus, which is itself a structural signal: the designers are testing a single reasoning thread across multiple items rather than three independent questions glued to a common exhibit.

The official description frames the family as the ability to evaluate information presented in multiple formats and to recognise when different sources are consistent, when they conflict, and when one source is simply irrelevant to the question being asked. That framing sounds generic, but in practice the family tests three very specific reasoning moves. The first is reconciliation: a number on Tab 2 contradicts a number on Tab 3, and the candidate has to decide which source is authoritative for the current question. The second is exclusion: Tab 1 and Tab 2 jointly support an inference, but Tab 3 introduces a constraint that rules it out, and the candidate has to spot the constraint. The third is sufficiency framing: the question itself supplies additional information in the stem, and the candidate has to judge which tab actually contains the data needed to apply that information.

For most candidates reading this, the family is also the most time-expensive item type in the section, because the cost of re-reading is high. Each tab is roughly 80 to 150 words of dense prose, often with a small embedded table, and the temptation is to read all three tabs in full before looking at the first question. In my experience this is the single most common pacing mistake on the GMAT Focus. The first tab read takes about 60 seconds, the second tab another 60 seconds, the third tab another 60 seconds, and by the time the candidate turns to the first question, working memory is saturated and the answer choices look interchangeable. The pacing fix is to read the first question stem before opening the second tab, which I will return to in the triage section below.

There is one further structural fact worth internalising. Multi-Source Reasoning sets are the only GMAT Focus items where the second or third question in a set will sometimes refer back to the candidate's answer to the first question. This is rare, but it is allowed, and it changes how the items should be sequenced during review. If a candidate runs out of time on item 2 of a 3-item MSR set, they should mark item 3 for review rather than abandoning it, because item 3 can in some cases be solved from the same working memory that was loaded for item 2. That structural quirk is one of the reasons the family rewards a steady, note-driven pace over a sprint-and-pray approach.

The four-minute budget: how to triage the tabs before reading them

A typical 3-item Multi-Source Reasoning set should be completed in roughly 4 minutes total, which works out to about 80 seconds per item once the triage is done. The triage itself should take 25 to 35 seconds and is the highest-leverage 30 seconds in the entire set. The goal of the triage is not to understand the data; the goal is to map the terrain so that the subsequent reading has a destination.

Step one of the triage is to read the first question stem only, not the answer choices. The stem will almost always name the variable the question cares about, whether the question is asking for a number, an inference, or an evaluation of a plan, and what the unit of analysis is. A stem that opens with "Based on the data shown" is a different reading task from a stem that opens with "Which of the following would most strengthen" or "If the company were to discontinue…". The verb in the first sentence of the stem determines the reading strategy for the tabs.

Step two is to open each of the three tabs in turn, but only for as long as it takes to read the tab's title, its first sentence, and any column headers or row labels. This pass should take 5 to 8 seconds per tab and produces a three-line mental summary: Tab 1 is a table of regional revenue, Tab 2 is a memo about cost allocations, Tab 3 is a footnote about a methodology change. The candidate now knows which tab is likely to contain the primary data, which tab is likely to contain the methodology constraint, and which tab is most likely to be a distractor.

Step three is to look at the second and third question stems in order, again without the answer choices, and to mark mentally which tabs each one is likely to need. Items 2 and 3 of an MSR set very often re-use the same Tab 1 and Tab 2, and introduce Tab 3 only on the third item. Recognising that pattern in the triage phase saves a full tab re-read on item 3, which is the difference between finishing the set in 3:30 and finishing it in 4:45.

Step four is the single line of notes that protects the rest of the set. The candidate should write, in roughly 10 seconds, a one-line header above each tab reference: for example, "T1 = revenue 2022, T2 = cost memo, T3 = methodology footnote, applies to 2022 only". This is the only note-taking I recommend on the GMAT Focus, because the on-screen scratchpad is finite and other item families do not benefit from persistent notes. On MSR sets, however, this single line is what allows the candidate to skip the second tab re-read on item 2 and the third tab re-read on item 3. Without it, the candidate re-reads; with it, the candidate reasons.

The four conflict patterns that decide the answer on a hard item

Hard Multi-Source Reasoning items are hard for a structural reason: they are engineered so that the answer choices look like they could be supported by a single tab, and the candidate has to read the question stem carefully enough to realise that the support actually lives in a different tab. The four patterns below account for the overwhelming majority of wrong answers on items 1, 2, and 3 of an MSR set, and recognising them by shape is faster than reasoning from first principles every time.

The first pattern is the silent-unit mismatch. Tab 1 reports revenue in millions of dollars, Tab 2 reports cost in thousands of dollars, and the question stem asks for a margin. A candidate who reads Tab 1 and answers from memory will pick a number that is off by a factor of 1,000. The fix is mechanical: convert units on the on-screen scratchpad the first time the value is read, and write the converted value next to the original. Once the conversion is on the scratchpad, it is safe to forget; without it, the candidate carries a unit error for the remaining two items.

The second pattern is the temporal trap. Tab 1 reports data for fiscal year 2022, Tab 2 reports projections for fiscal year 2023, and Tab 3 is a footnote stating that the company changed its accounting policy in March 2023. A question stem that asks "what was the company's revenue in 2022" can be answered from Tab 1 alone; a stem that asks "what will the company's revenue be in 2023" requires Tab 2, and only Tab 3 tells the candidate whether Tab 2's projection is comparable to Tab 1's historical number. Most wrong answers on hard MSR items come from answering the temporal question that was not asked.

The third pattern is the scope expansion. Tab 1 reports data for the North American segment, Tab 2 reports data for the European segment, and the question stem asks about the company as a whole. A candidate who reads Tab 1 and Tab 2 sequentially will often treat the two as additive, when in fact the question requires the candidate to recognise that the data does not cover Asia-Pacific at all. The answer choices in a scope-expansion item are designed to exploit exactly this confusion: two choices will be the sum of the two reported segments, one choice will be the sum of the two reported segments minus a clearly labelled double-count, and one choice will explicitly state that the data is insufficient. The correct answer is almost always the explicit insufficiency answer, because the GMAT Focus scores cross-source reasoning, not the candidate's willingness to add two imperfect numbers together.

The fourth pattern is the authority inversion. Tab 1 is a press release, Tab 2 is an internal memo, and Tab 3 is an audited financial statement. The question stem asks which source should be treated as authoritative for a specific claim. The hard version of this item is the one where the press release and the internal memo agree, and the audited statement disagrees, and the candidate has to recognise that the audited statement overrides both. The fix is to treat the most procedurally rigorous source as authoritative by default: audited statements, regulatory filings, and footnotes with specific dates outrank press releases, marketing decks, and unlabelled bar charts. The candidate who reasons from source quality rather than source recency is the candidate who answers the authority-inversion item correctly.

Note-taking that survives the second item

The on-screen scratchpad on the GMAT Focus is shared across the entire Data Insights section, and it does not persist between questions. Anything the candidate writes while reading Tab 1 is therefore at risk of being overwritten by a calculation performed for an unrelated Graphics Interpretation item three questions later. This is the structural reason MSR notes have to be more disciplined than other section notes.

The discipline I recommend is a three-line header per tab, written in a fixed corner of the scratchpad, that names the tab, the unit, and the time period. For a 3-item MSR set, that is nine lines of notes plus any specific numbers the candidate extracts from the question. Nine lines is enough to fit on a single screen and is short enough that the candidate can re-orient in 10 seconds at the start of item 2 and again at the start of item 3. Anything longer than nine lines, and the notes themselves become a re-reading burden.

For most candidates reading this, the failure mode is the opposite of what they expect. The failure mode is not taking too few notes; it is taking too many. A candidate who copies the entire Tab 1 table onto the scratchpad has, in effect, turned the scratchpad into a second copy of the tab, and now has to read both the tab and the scratchpad on item 2. The candidate who writes only the column headers, the units, and the time period has a single source of truth, and that source is the scratchpad. The tab itself becomes a reference to be opened only when the question forces a number lookup.

One tactical note on numerical values: candidates should write the number down the first time they read it, but they should write it in the form the question will ask for. If the tab reports revenue as 1,200 and the question asks for revenue in millions, the candidate should write 1.2, not 1,200. Pre-converting the number onto the scratchpad removes one of the four conflict patterns described above (the silent-unit mismatch) without any additional reading time. It also makes the calculation step on item 2 and item 3 a single arithmetic operation rather than a two-step conversion plus arithmetic, which is faster and less error-prone.

Comparison table: the four item families in GMAT Focus Data Insights

The Multi-Source Reasoning family is one of four item families in the Data Insights section, and the way it differs from the other three is the reason it deserves its own preparation track. The table below summarises the structural differences. The rows are written for a candidate who is triaging a section under time pressure, not for a curriculum designer, and the focus is on what the candidate has to do differently at the keyboard.

Feature	Data Sufficiency	Table Analysis	Graphics Interpretation	Multi-Source Reasoning
Items per stimulus	1	1	1	2 to 3
Primary skill tested	Categorise the question, then judge sufficiency	Sort and filter a table to isolate a subset	Read axis labels, then read a value	Merge three tabs into a single mental model
Re-reading cost between items	None (one item only)	None (one item only)	None (one item only)	High — items 2 and 3 re-use earlier tabs
Note-taking value	Low	Medium	Low	High — protects item 2 and item 3
Typical time budget	2 minutes	2 minutes	2 minutes	4 minutes for a 3-item set
Most common wrong-answer pattern	Calculation when categorisation is enough	Sorting by the wrong column	Reading the wrong axis	Answering from a single tab when two are needed

Two rows in this table drive the rest of the preparation strategy. The first is the re-reading cost: MSR is the only family where the items are linked, which means the candidate who masters tab triage and scratchpad notes gets a compounding time saving across the set. The second is the most common wrong-answer pattern: candidates who treat MSR as a sequence of three independent reads will answer the first item correctly and the second or third item incorrectly, because the second and third items are explicitly engineered to require cross-source reasoning.

Worked example: a 3-item MSR set, end to end

The fastest way to internalise the method is to walk a full set from the first triage line to the third answer choice. The example below is a representative 3-item set, simplified for exposition but structurally faithful to the live exam. The tabs are described in prose; the candidate would see them as three small panels above the question stem.

Tab 1 is a small table showing quarterly revenue for a software company across four regions, in millions of dollars, for fiscal years 2022 and 2023. Tab 2 is a one-paragraph memo from the CFO stating that the company reclassified its services revenue starting in Q3 2023, and that figures prior to that date have not been restated. Tab 3 is a footnote from the annual report stating that the company's 2023 figures include a one-time gain from the divestiture of a European subsidiary, and that the gain is reported as a separate line item.

Item 1 asks: "Based on the data shown, in which quarter did the company first report revenue exceeding 200 million dollars across all regions combined?" The triage tells the candidate that this is a numerical lookup question, that the answer is on Tab 1, and that the unit is millions of dollars. The candidate adds the four regional columns for each quarter, scans for the first quarter where the sum exceeds 200, and selects the answer choice corresponding to that quarter. The second and third tabs are not needed for item 1, and the candidate should make a mental note that they are likely to be needed for items 2 and 3.

Item 2 asks: "If the CFO's reclassification had been applied retroactively to 2022, would the company's reported 2022 revenue have been higher, lower, or the same?" This is the authority-inversion pattern, with a temporal twist. The candidate has to recognise that Tab 2 explicitly states that 2022 figures have not been restated, which means the reclassification cannot be applied retroactively, which means the answer is "the same, because the reclassification only affects 2023". The wrong answer choices are designed to lure the candidate into a calculation: they offer a specific 2022 number that would be the answer if the reclassification had been applied. The candidate who reads the question stem first recognises that the verb is "would have been" and that the constraint in Tab 2 forecloses the calculation.

Item 3 asks: "Which of the following statements about the company's 2023 revenue is most supported by the data shown?" This is the cross-source inference question, and it requires all three tabs. The candidate has to recognise that Tab 1's 2023 figures include a one-time gain (Tab 3), that the gain is reported as a separate line item (Tab 3), and that the reclassification (Tab 2) means 2022 and 2023 are not directly comparable. The correct answer is the choice that explicitly acknowledges the non-comparability; the wrong answer choices are two plausible-looking comparisons (one for 2023 only, one for 2022 versus 2023) that ignore Tab 3's line-item disclosure. The candidate's scratchpad note from the triage phase — "T2 = reclassification, T3 = one-time gain as separate line item" — is what allows the candidate to answer item 3 without re-reading the tabs in full.

The full time budget for this worked set, with the method applied, is approximately 3 minutes 45 seconds: 30 seconds of triage, 75 seconds to read Tab 1 and answer item 1, 60 seconds to re-open Tab 2 and answer item 2, 60 seconds to re-open Tab 3 and answer item 3, plus the time spent on the on-screen scratchpad. Without the method, the same set typically takes 5 minutes 15 seconds, because the candidate re-reads all three tabs on every item. The 1 minute 30 second saving compounds across a section that contains more than one MSR set, and is the difference between finishing the Data Insights section with a buffer and finishing it under pressure.

Common pitfalls and how to avoid them

The Multi-Source Reasoning family has a small number of recurring failure modes, and most of them are addressable with a single tactical change. The list below is the senior-tutor shortlist, in the order in which they tend to appear on the live exam.

Reading all three tabs before the first question. The candidate who reads all three tabs in full before turning to the first question has spent 3 minutes reading and 0 minutes answering. The fix is the triage described above: read the first stem, skim the three tab headers, then open the tab that the first stem actually needs.
Answering item 2 from a single tab. The candidate who re-reads only Tab 1 on item 2 and ignores Tab 2 will pick an answer choice that is supported by Tab 1 alone, and the GMAT Focus has engineered item 2 to require Tab 2. The fix is the scratchpad note: if the triage note said "T2 needed for item 2", the candidate opens Tab 2 before committing to an answer.
Forgetting the reclassification between items. The candidate who answers item 2 correctly but then forgets, on item 3, that the reclassification in Tab 2 means 2022 and 2023 are not comparable, will pick a comparison-style answer choice on item 3. The fix is to re-read the scratchpad note at the start of item 3, which is exactly what the note is for.
Running out of time on item 3 because item 2 overran. The candidate who spends 2 minutes 30 seconds on item 2 has, in effect, lost item 3. The fix is a hard 90-second cap on item 2: if item 2 is not answered in 90 seconds, the candidate marks it, moves to item 3, and returns to item 2 only if time remains. The 90-second cap is the single most effective pacing rule on MSR sets.
Treating the third tab as a footnote. Tab 3 in a 3-item MSR set is almost always the source of the hardest inference, because the test designer has placed the most disqualifying constraint on the tab the candidate is least likely to read. The fix is to assume, on every MSR set, that Tab 3 contains the constraint, and to read it in full before answering item 3.

Preparation strategy: how to drill MSR without burning the question bank

The standard preparation strategy for the GMAT Focus Data Insights section is to drill by item family rather than by question, because the cross-source reasoning skill is family-specific and does not transfer cleanly from Data Sufficiency to Multi-Source Reasoning. The drill that works for MSR, in my experience, is the four-pass method, which trains triage, tab-reading, cross-source inference, and pacing in sequence rather than all at once.

Pass one is triage only. The candidate opens an MSR set, reads the first stem, skims the three tab headers, and closes the set without answering. The goal is to internalise the 25 to 35 second budget and to produce the nine-line scratchpad note. Ten sets, done back to back, will take roughly 6 minutes and will produce a measurable improvement in tab-mapping speed.

Pass two is full execution under a 5-minute cap. The candidate takes a fresh set, applies the full method, and stops at 5 minutes even if the third item is incomplete. The goal is to expose the failure modes above in a low-stakes setting. The candidate should review only the items that ran over the 5-minute cap, and should diagnose the overrun as either a triage failure (spent too long reading tabs) or an inference failure (spent too long on the answer choices).

Pass three is the 3-item set, fully timed, with the hard version of the family only. The candidate restricts the drill to MSR sets that include an authority-inversion item or a scope-expansion item, and applies the 4-minute target. The goal is to build fluency on the items that account for the most wrong answers on the live exam.

Pass four is the integration drill. The candidate takes a full-length Data Insights section, treats the MSR sets as ordinary items inside the section, and reviews only the items that broke the pacing budget. The goal is to confirm that the MSR method does not interfere with the candidate's pacing on the other three item families, which is the preparation-stage mistake I see most often. A candidate who over-prepares for MSR can lose 30 seconds per Data Sufficiency item, because they have trained themselves to re-read the stem three times when a single read is enough.

Conclusion and next steps

Multi-Source Reasoning is the only item family in the GMAT Focus Data Insights section that explicitly tests the candidate's ability to compare multiple data sources inside a single question set. The skill is family-specific, it does not transfer from Data Sufficiency or Graphics Interpretation, and the candidates who score 705+ on the GMAT Focus composite are the candidates who have drilled tab triage, scratchpad note-taking, and the four conflict patterns described above until the method runs on autopilot. The preparation track for this item family is short, it is highly specific, and it pays off disproportionately in the scaled Data Insights score. TestPrep Europe's diagnostic assessment is a natural starting point for candidates building a sharper preparation plan for the cross-source reasoning sub-family of the GMAT Focus Data Insights section.

Frequently asked questions

How many Multi-Source Reasoning sets appear on the GMAT Focus Data Insights section?

The GMAT Focus Data Insights section contains a small number of Multi-Source Reasoning sets, each with two or three linked items. The exact count is not fixed across forms, but in practice a candidate will see more than one set, which is why the cross-source reasoning skill is worth drilling as its own preparation track rather than as a one-off item type.

Should I read all three tabs before answering the first question on an MSR set?

No. The standard recommendation from senior tutors is to read the first question stem first, then skim the three tab headers to map the terrain, then open only the tab the first question actually needs. Reading all three tabs in full before turning to item 1 is the single most common pacing mistake on Multi-Source Reasoning, because it loads working memory with information the first question does not require.

What should I write on the scratchpad during an MSR set?

A three-line header per tab is enough, naming the tab, the unit of measurement, and the time period the data covers. Specific numbers should be written in the form the question will ask for, so that the silent-unit mismatch pattern does not catch you on items 2 or 3. The scratchpad is shared across the entire Data Insights section, so MSR notes should be kept to a single screen and reviewed at the start of each new MSR item.

How long should I spend on a 3-item Multi-Source Reasoning set?

A 3-item set should be completed in roughly 4 minutes total under timed conditions, which works out to about 80 seconds per item after the triage. The triage itself should take 25 to 35 seconds. If item 2 has not been answered within 90 seconds, the recommended tactic is to mark it, move to item 3, and return to item 2 only if time remains at the end of the set.

Does Multi-Source Reasoning appear on the GMAT Focus in a different form than on the classic GMAT?

The Multi-Source Reasoning family on the GMAT Focus is the structural descendant of the same family on the classic GMAT, with adjustments to fit the shorter Data Insights section. The cross-source reasoning skill, the three-tab layout, and the linked-item structure are preserved. Candidates who prepared for the classic GMAT MSR family will recognise the question type, but should re-drill the pacing because the GMAT Focus section length is shorter and the time budget per set is tighter.

4 conflict patterns that decide your GMAT Focus Multi-Source Reasoning answer