GMAT Critical Reasoning: how to find the hidden assumption.

Most students treat Critical Reasoning as a reading test. It isn't. It's a logic test wearing a paragraph as a costume. The questions feel different from each other — one asks you to strengthen, one to weaken, one to find the flaw — but underneath, they reward the same single move: seeing the gap between what the author proved and what the author concluded, then naming the assumption that papers over that gap. Get fluent at finding the assumption and the rest of the section stops being a collection of question types and becomes one skill applied four ways.

An argument never proves its conclusion outright. It always leans on something it didn't say. Your job is to find the thing it's leaning on.

The anatomy of an argument

Every Critical Reasoning argument has two parts that matter and one that distracts. The parts that matter are premises (the evidence the author offers) and the conclusion(the claim the author wants you to accept). Between them sits a logical gap, because the premises almost never fully cover the conclusion. The distracting part is background — context, history, the opposing view — that's there to set the scene, not to do logical work.

Consider a short argument. A regional bakery switched from a national flour supplier to a local mill last quarter. Its profit margin rose two points that quarter. So switching to the local mill made the bakery more profitable. The premises are the switch and the margin increase. The conclusion is the causal claim that the switch causedthe increase. The gap is enormous: a margin can rise for a dozen reasons that have nothing to do with flour. The author has quietly stepped from “two things happened together” to “one caused the other,” and that step is exactly where the assumption lives.

How to isolate the conclusion first

You can't find the gap until you know which claim is the conclusion. This is the step students rush, and it sinks them. Two reliable methods:

1. Hunt for the markers. Words like therefore, so, thus, hence, and clearlyusually sit right in front of the conclusion. But don't trust them blindly — the GMAT writes arguments where the conclusion comes first and the evidence follows, with no marker at all.
2. Run the why test.Take each candidate claim and ask, “Why does the author believe this?” If the answer is “because of the other statements,” you've found the conclusion. The conclusion is the claim that everything else is offered to support; a premise is a claim offered in support.

A clean way to lock it in: phrase the argument as “X, therefore Y.”Y is the conclusion, X is the premise set. If reversing them (“Y, therefore X”) reads as nonsense, you got the direction right. The conclusion is the main point being argued for — not necessarily the last sentence, and not the most dramatic sentence.

What an assumption actually is

An assumption on the GMAT is not a wild guess or a vibe. It has a precise definition: an unstated statement the argument requires in order to hold — something the author must believe but never wrote down. It's the missing plank in the bridge from premise to conclusion. If you removed it, the bridge would collapse.

Back to the bakery. For “the switch made us more profitable” to follow from “we switched and margins rose,” the author has to be assuming that nothing elsemeaningfully drove the margin up that quarter — not a price increase, not a fall in some other cost, not a seasonal sales spike. That “nothing else” idea is never stated. The author needs it anyway. It is the assumption.

Two traits separate a real assumption from a plausible-sounding distractor. First, it's unstated— if it's written in the passage, it's a premise, not an assumption. Second, it's required — the argument genuinely depends on it, rather than merely being helped by it. The second trait is where most wrong answers die, and it's exactly what the Negation Test measures.

The Negation Test: your core tool

Here is the single most useful technique in Critical Reasoning. When you have a candidate assumption and you're not sure whether the argument truly needs it, negate the candidate— flip it to its opposite — and ask what happens to the conclusion.

• If the negated statement destroys the argument— the conclusion can no longer stand — then the original statement was necessary. It's a real assumption.
• If the negated statement leaves the argument intact— the conclusion still survives — then the original wasn't required. It's a distractor.

The logic is clean: a necessary condition is one whose absence breaks the thing. If knocking out a statement breaks the argument, the statement was load-bearing. Walk it through end to end on a fresh example.

A worked example, start to finish

A software company found that employees who use its new internal chat tool close their support tickets 20 percent faster than employees who don't. To raise overall ticket speed, the company plans to require everyone to use the chat tool.

First, isolate the conclusion. Run the why test: why does the company believe requiring the tool will raise speed? Because the tool-users are faster. So the conclusion is “requiring the tool will raise overall ticket speed,” and the premise is the 20 percent gap.

Now spot the gap. The premise compares people who chose to use the tool against people who chose not to. The conclusion is about forcingeveryone to use it. The author is assuming that the speed difference comes from the tool, not from the people — that the fast employees aren't simply the more capable ones who would be fast with any tool. Candidate assumption: “The faster ticket-closing is caused by the chat tool, not by pre-existing traits of the employees who adopted it voluntarily.”

Apply the Negation Test. Negate it: the faster ticket-closing is NOT caused by the tool — it comes from the voluntary adopters being the more capable employees. If that's true, then forcing the tool on everyone does nothing for speed; you'd just be handing a tool to people whose performance was never about the tool. The conclusion collapses. The negation destroys the argument, so the assumption was necessary. That's the answer an Assumption question is hunting for.

Negate the candidate. If the opposite kills the conclusion, you found the assumption. If the conclusion shrugs it off, you found a trap.

Necessary vs sufficient assumptions

This distinction trips up strong students, so be precise about it. A necessary assumption is one the argument musthave to survive — remove it and the argument fails. A sufficient assumption is one that, if added, would be enoughto guarantee the conclusion — but the argument might not strictly require it.

GMAT Assumption questions almost always want the necessaryone. That's why the Negation Test works: it's a test for necessity, not sufficiency. Practically, this means the right answer is often modest and narrow — “the two groups are otherwise comparable,” “no major alternative cause was at work” — rather than sweeping. Beware answer choices that prove too much. A choice that single-handedly guarantees the conclusion is frequently a sufficient-assumption decoy, not the necessary assumption the question asks for. When two answers both seem to help, the necessary one is usually the weaker-sounding, more cautious statement.

The assumption families the GMAT reuses

The test recycles a small set of logical gaps. Once you can name the family on sight, you find the assumption in seconds, because you already know the shape of what's missing.

1. No alternative cause

The argument sees a correlation and concludes causation. The assumption is that nothing else explains the result. This is the most common family on the test. Any time an argument says two things moved together and therefore one caused the other, the buried assumption is “no other cause was responsible, and the causation doesn't run the other way.”

2. The sample is representative

The argument studies a group and generalizes to a larger population. The assumption is that the group studied fairly represents the whole. A consultancy surveyed its three most profitable clients and concluded that all its clients value speed over price.The buried assumption: those three are typical of the full client base — which, being the most profitable, they very likely are not.

3. The analogy holds

The argument says one case worked, so a parallel case will too. The assumption is that the two situations are alike in the ways that matter. A pricing model that worked for the company's urban stores will work for its rural ones. The hidden claim is that urban and rural markets are similar where it counts.

4. A term keeps one meaning throughout

A word shifts meaning between premise and conclusion, and the argument assumes it didn't. The team is more “efficient” now — so it produces more output per hour.If “efficient” in the evidence meant “fewer meetings” rather than “more output,” the argument equivocates. The assumption is that the term means the same thing both times.

5. What is true of the part is true of the whole

The argument takes something true of pieces and concludes it's true of the combination, or vice versa. Every department cut its own costs, so the company's total costs fell.Not necessarily — shared overhead or cross-department transfers can rise even as each unit economizes. The assumption is that the parts simply add up to the whole.

6. Nothing will change in the future

The argument projects a past pattern forward and assumes conditions stay put. Demand grew 10 percent a year for three years, so it will grow 10 percent next year too. The assumption is that no new competitor, regulation, or shift in taste interrupts the trend.

7. No unintended side effect

A plan is proposed to achieve a goal, and the argument assumes the plan won't trigger a consequence that defeats it. Cutting the warranty period will lower costs, so profits will rise.The assumption is that the cut won't drive away enough customers to wipe out the savings. Plan-based arguments almost always hide a no-backfire assumption.

Why this one skill unlocks four question types

Here's the payoff. The four most common Critical Reasoning question types are not four separate skills. They are four things you do to the same assumption. Find the assumption first, and each question type tells you what to do with it.

• Strengthen— the correct answer affirms or supports the assumption, closing the gap. For the chat-tool argument, a strengthener confirms the two employee groups were comparable to begin with.
• Weaken— the correct answer attacks the assumption, reopening the gap. A weakener shows the voluntary adopters were already the top performers, so the tool wasn't the cause.
• Flaw— the correct answer names the missing link in the abstract: “the argument assumes a correlation proves causation” or “it treats a self-selected group as representative.” A flaw answer is just the assumption family stated as a criticism.
• Evaluate— the correct answer poses the question whose answer would tell you whether the assumption holds: “Were the two groups equally skilled before the tool was introduced?” If a yes-versus-no answer swings the argument's strength, you've found the evaluator.

That's the whole game. One core move — locate the premise-to-conclusion gap and name what fills it — powers the majority of the section. You stop memorizing per-type tricks and start running the same diagnostic every time: find the conclusion, find the gap, name the assumption, then do whatever the question stem asks you to do with it.

How to drill this

1. Pre-phrase before reading the choices. After the stimulus, state the assumption in your own words first. Then go to the answers looking for a match. This stops the wrong answers from talking you out of the right one.
2. Negate every finalist.On Assumption questions, don't just pick the choice that sounds right — negate your top two and confirm that the right one's negation breaks the argument while the other's doesn't.
3. Tag the family.For every CR question you miss, note which assumption family was at play. Within a few weeks you'll see the test reuses the same handful, and you'll spot them before you finish reading.
4. Practice across types together.Mix Strengthen, Weaken, Flaw, and Evaluate in one set. Forcing your brain to find the assumption first — before the stem tells you what to do — builds the unified skill instead of four brittle ones.

The short version

Isolate the conclusion with the why test. Find the gap between it and the premises. Name the unstated bridge the argument needs — almost always one of the seven families. Confirm it with the Negation Test: flip it, and if the argument dies, it's the assumption. Remember the test wants necessary, not sufficient. Then point that one finding at whatever Strengthen, Weaken, Flaw, or Evaluate question is in front of you. Critical Reasoning isn't a reading section. It's one logical move you make over and over.

The platform

Zakarian GMAT's Verbal chapters teach the assumption skill as the backbone of Critical Reasoning, then drill it across all four question types in mixed sets so you build the unified move rather than four shaky ones. The error log's six-tag taxonomy lets you tag CR misses by family so you can see which assumption gaps keep catching you. The sample chapter is free if you want to see the teaching first.