V2 · The Question Types
V1 gave you one sentence: this concludes X, on the basis of Y, and it only works if Z. This chapter is what you do with Z once you have it.
There are eight question types. That sounds like eight things to learn, and it is not — it is one thing asked eight ways. The clearest way to see that is to take a single argument and put four different stems on it.
A regional hospital placed a triage nurse at its emergency entrance. In the following year, average waiting times fell by 22 percent. The hospital board has concluded that the triage nurse produced the reduction.
The gap, from V1: the argument reports that two things happened in sequence and concludes that one caused the other. It works only if nothing else explains the fall.
Now watch the same gap answer four different questions.
"The board’s conclusion depends on which of the following assumptions?"
You are being asked to state the gap. The answer will be something like: the reduction was not caused primarily by some factor unrelated to the triage nurse. Nothing new — just Z, written out.
"Which of the following, if true, most strengthens the board’s conclusion?"
You are being asked to close the gap. An answer might be: waiting times at the hospital’s three other emergency departments, which have no triage nurse, were unchanged over the same period. That removes the leading alternative.
"Which of the following, if true, most seriously undermines the conclusion?"
You are being asked to widen the gap. An answer might be: the hospital added four staff to the emergency rota in the same month. Now something else explains the fall.
"Which of the following would be most useful to determine in evaluating the conclusion?"
You are being asked what would test the gap. An answer might be: whether emergency admissions fell over the period. If they did, fewer patients explains the shorter waits; if they did not, the nurse looks more credible.
Four stems, four correct answers, one gap. The stem does not change what you are looking at. It changes what you are asked to do with it.
It means the fifteen seconds you spend on V1’s sentence does most of the work for every type in this chapter. It means that when you find yourself unable to evaluate the options, the problem is usually that you have not named the gap — not that you have forgotten a technique.
And it means the eight types are learnable in an afternoon, because seven of them are variations on a move you already know.
Three types do not fit the pattern cleanly, and this chapter treats them separately because they need it.
Inference inverts the format: the passage is a set of facts rather than an argument, and the correct answer must follow with certainty rather than plausibly. Strong language, which is usually right elsewhere, is usually wrong here.
Paradox gives you two facts that appear to conflict and asks what makes both true at once. There is no conclusion to attack.
Boldface asks about structure rather than content — what role each marked portion plays — which means knowing what the argument says is not sufficient.
By the end you will recognise the stem type in two seconds, know which of four moves it is asking for, and — crucially — know what the correct answer will look like before you read the options. That last part is what separates working through five options from choosing between them.
"The argument depends on which of the following assumptions?" · "Which of the following is an assumption required by the argument?" · "The conclusion is properly drawn if which of the following is assumed?"
What you are looking for: something the argument needs in order to work — not something that would help it, not something that sounds reasonable. Necessary, not merely supportive.
That distinction is the whole difficulty of this type, and there is a reliable test for it.
Take the answer choice, negate it, and ask whether the argument collapses. If negating it destroys the argument, it was necessary — that is your answer. If negating it leaves the argument standing, it was never required, however sensible it sounds.
A regional dairy switched its herd to a new feed blend, and milk yield per cow rose 8 percent over the following year. The manager plans to introduce the same blend at a second farm, expecting a similar increase.
Candidate assumption: "Factors other than the feed blend were not primarily responsible for the increase at the first farm." Negate it: factors other than the feed were primarily responsible. Now the 8 percent tells you nothing about the feed, and the plan for the second farm has no support at all. The argument collapses. Necessary.
Compare with: "The feed blend costs no more than the feed it replaces." Negate it: the blend costs more. The plan is now more expensive — but the manager's claim was about yield, not cost, and the yield reasoning is untouched. The argument survives. Not an assumption.
The most common wrong-answer shape is exactly that second one: something the manager would probably want to be true, or that would make the plan wiser, but that the reasoning does not require. Assumption questions are about logical necessity, not about good business judgement.
"Which of the following, if true, most strengthens the argument?" · "…provides the most support for the conclusion?"
What you are looking for: something that closes the gap — usually by ruling out an alternative explanation, by confirming that the sample transfers, or by establishing that conditions have not changed.
Note the "if true". You do not evaluate whether the option is plausible. You accept it and ask what it does to the argument.
Archaeologists excavating a coastal settlement found fish bones in nearly every dwelling but almost no farming tools. They concluded the inhabitants relied primarily on fishing rather than agriculture.
The gap: absence of tools is being read as absence of farming. That step needs the tools to have survived if they existed.
An option like "Similar inland settlements from the same period contain abundant farming tools and few fish bones" strengthens it — and notice how. It establishes that this soil and this period do preserve farming tools when they were used. The absence here becomes meaningful.
- Restating a premise. An option that says the same thing the argument already told you adds nothing. If the argument said fish bones were found everywhere, an option confirming that fish were eaten is not new support.
- Supporting the topic rather than the reasoning. An option establishing that the settlement was important, or long-occupied, or coastal, is about the subject and not about the inference.
"Which of the following, if true, most seriously undermines the argument?" · "…casts the most doubt on the conclusion?"
What you are looking for: something that breaks the link between premise and conclusion. Not something that contradicts a premise — you accept those — and not something that makes the topic look bad.
For a causal argument, which is most of them, the strongest weakener is almost always one of three:
| Shape | What it does |
|---|---|
| Alternative cause | Something else explains the effect |
| Effect without cause | The result occurred where the supposed cause did not |
| Reversed direction | The effect is causing the "cause" |
A ferry company introduced an online booking system, and passenger numbers rose 12 percent the following year. The company credits the booking system.
"A competing ferry route on the same crossing ceased operating three months before the booking system launched" is an alternative cause, and a strong one: the displaced passengers had to go somewhere, and the timing precedes the system.
"Which of the following would be most useful to determine in evaluating the argument?" · "The answer to which of the following questions would be most helpful in assessing…?"
What you are looking for: a question whose two possible answers push the argument in opposite directions. That is the test, and it is mechanical.
Take the option, imagine the answer is yes, and ask what happens to the argument. Then imagine the answer is no. If the argument is stronger one way and weaker the other, that is your answer. If both answers leave it roughly where it was, the option is irrelevant however interesting it sounds.
A school district replaced fried cafeteria offerings with baked alternatives, and food waste rose 30 percent. The nutrition director concluded that students dislike the baked options.
Option: "Whether the total quantity of food served increased when the menu changed."
If yes — more food was served — then 30 percent more waste may be proportional and says nothing about preference. The argument weakens badly.
If no — the same quantity was served — then the extra waste is genuinely about what students left on the tray. The argument strengthens. Opposite directions. That is the answer.
This type is the most mechanical of the eight, and candidates under-use the technique because it feels like extra work. It is faster than reasoning about which option "seems most relevant", and it does not depend on judgement.
Practise these four types 16 structure drills · 32 full items · a diagnosis across all four →"Which of the following must be true?" · "Which of the following can be properly inferred?" · "The statements above, if true, most strongly support which of the following?"
This type is different from the other seven, and it catches people badly.
There is no argument here. There is no conclusion to attack or defend. You are given statements and asked what follows necessarily from them. The direction is reversed: instead of finding what the argument needs, you find what the statements guarantee.
The correct answer often feels underwhelming. It restates a small, safe consequence of what you were given. Candidates skip past it looking for something more insightful, and choose an option that goes one step beyond the text — which is exactly the trap.
Every glass container collected last year was clear, green or brown. All the green containers came from commercial premises. Some containers from commercial premises were brown.
"Some brown containers were collected" must be true — the third statement guarantees it. It feels almost too obvious.
Whereas "No clear containers came from commercial premises" sounds like a sharper observation and is not supported by anything: the statements say where the green ones came from, not where the clear ones did.
For each option, ask "could this be false while everything I was told stays true?" If you can imagine one case where it fails, it is out. Use the quantifier table from V1.5 — some means at least one, and it never tells you about proportion.
"Which of the following, if true, best explains the discrepancy?" · "…most helps to resolve the apparent paradox?"
What you are looking for: something that makes two surprising facts compatible. Both facts stay true — your job is to find the missing circumstance under which they can coexist.
First, state the tension explicitly: "You would expect X, but Y happened." Candidates who skip this step end up choosing an option that explains only one of the two halves.
A wind farm installed sensors that shut down turbines when bat activity is high, expecting a significant drop in annual output. Output fell by less than 1 percent.
The tension: shutdowns should cost output, and barely did.
"Bat activity is highest on nights with low wind speed, when the turbines generate little electricity anyway" resolves it completely. The shutdowns happened — they simply landed on hours that were worth almost nothing.
Wrong answers on this type typically address only one side. An option about how much the sensors cost, or about bat populations, leaves the discrepancy exactly where it was.
"In the argument given, the two boldfaced portions play which of the following roles?"
This is the type candidates fear most, and the fear is misplaced. It is mechanical — it just has two things to get right instead of one, which is why the five options often share a first half or a second half.
- Find the argument's own conclusion. Not the position it describes — its own. The V1.2 "because" test does this.
- For each bold portion, ask what it is: the conclusion, a premise supporting the conclusion, a position the argument opposes, or evidence against that opposed position?
- Only then read the options, and eliminate on the first half before reading the second. If the first half is wrong, the option is dead regardless of how good the second half sounds.
That third step is where the time is saved. Options in this type are built so that each half is plausible on its own; reading them as whole sentences means evaluating ten claims instead of five.
Newspapers that moved to online-only publication have generally seen subscription revenue fall. A publisher considering the move argues its case is different because its readership skews younger and more digitally engaged. However, the papers with the steepest declines were precisely those with the youngest readerships.
The first bold is a generalisation the argument accepts and builds on. The publisher offers an objection to it. The second bold is evidence that the objection does not hold.
Note what neither portion is: neither is the conclusion. The argument's conclusion — that the publisher's case is not different — is never stated outright. That is common in this type, and it is why step 1 has to happen before you look at the options.
This is not a ninth question type. It is the reasoning flaw that sits underneath more items than any other, appearing in assumption, strengthen, weaken and evaluate questions alike — which is why it gets its own section.
The pattern: two things are observed together, and the argument concludes that one causes the other. Four ways that can be wrong, and you should be able to name all four on sight:
| Name | The alternative |
|---|---|
| Reverse causation | Y causes X, not the other way round |
| Third factor | Z causes both X and Y |
| Coincidence | The association is chance, or the sample is small |
| Selection | The group was not assigned — it selected itself |
That last one is the most under-recognised and the most common in this format. Whenever a premise describes people who chose something — who read often, who took the stairs, who attended the workshop — the group differs from everyone else in ways beyond the choice itself.
A study of 4,000 adults found that those who read for pleasure at least four times a week had markedly better memory scores. A commentator concluded that reading for pleasure improves memory.
"People with better memory find reading more enjoyable and therefore read more often" is reverse causation, and it accounts for the entire correlation without any effect of reading on memory.
Weaken — supply one of the four alternatives. Strengthen — rule one out. Assumption — the argument assumes none of them applies. Evaluate — ask a question that would establish whether one does.
A causal claim can fail in exactly these four ways, so the four names are a complete checklist rather than a set of hints. Running them takes about ten seconds and converts an open question — "what is wrong with this?" — into a lookup. That matters most under time pressure, when open questions are the ones you answer badly.
Every error this chapter has described, in one place. Nine traps across eight question types — open each to see the shape it takes and the defence against it.
Each item names its type, so you can practise the method rather than the identification. On the exam the stem tells you anyway. Name the gap in your own words before reading the options — that is the habit the whole chapter rests on.
The manager's plan depends on which of the following assumptions?
Negate it: other factors were primarily responsible. Then the 8 percent says nothing about the feed, and the plan for the second farm has no support. Collapses — necessary.
B is the classic sufficient-not-necessary trap: cost matters to a business decision and is irrelevant to a claim about yield. Negate it and the yield reasoning is untouched. A is a plausible-sounding transfer condition the argument does not require — a different breed might respond similarly, and the manager expects "a similar increase", not an identical one. C is a value judgement. E is about the first farm's future, not the second farm's prospects.
Which of the following, if true, most strengthens the archaeologists' conclusion?
The gap is that absence of tools is being read as absence of farming, which requires the tools to have survived if they existed. C establishes exactly that: the same period and comparable conditions do preserve farming tools elsewhere, so their absence here is meaningful.
A and B both weaken — each supplies a reason the evidence might be misleading. B is the sharper of the two and is the option most often chosen by candidates reading quickly. D is about duration. E raises a possibility that complicates the conclusion rather than supporting it.
Which of the following, if true, most seriously undermines the company's explanation?
An alternative cause, and a strong one: a competitor leaving the route sends its passengers somewhere, and the timing precedes the booking system.
A attacks the project's cost and leaves the causal claim intact — the attack-the-subject trap from V2.3. C is mild support. D is irrelevant to passenger numbers. E slightly strengthens, by ruling out a pre-existing upward trend.
The answer to which of the following questions would be most useful in evaluating the director's conclusion?
Run the two-way test. If the total quantity served increased, then 30 percent more waste may be simply proportional and says nothing about preference — the argument collapses. If it stayed the same, the extra waste is genuinely food left on trays — the argument strengthens. Opposite directions.
A is about cost, not preference. B would be useful if the question were whether the recipes were executed well, but a badly cooked baked option still means students left it. C and E are external comparisons that do not bear on what happened here.
If the statements above are true, which of the following must also be true?
The third statement says some containers from commercial premises were brown, which guarantees at least one brown container exists. It feels almost too small to be the answer, which is the V2.5 point exactly.
A sounds like a sharper observation and is unsupported — we were told where the green ones came from, not where the clear ones did. B is contradicted by the third statement. D introduces a proportion nothing establishes; some never tells you about quantity. E makes a claim about residential premises, which the statements never characterise.
Which of the following, if true, best explains this result?
Resolves both halves at once: the shutdowns did occur, and they landed on hours that generated almost nothing. Expectation and outcome become compatible with no contradiction.
A explains seasonal variation but not why the annual total barely moved. C is about cost. D is a future plan. E might explain fewer shutdowns, but if bat activity had declined enough to matter, the operator's expectation of frequent shutdowns would itself need explaining — it addresses at most one side.
In the argument given, the two boldfaced portions play which of the following roles?
Work step 1 first: the argument's own conclusion is that the publisher's case is not different — and it is never stated outright. Neither bold portion is the conclusion, which eliminates B and C on their first halves alone.
The first bold is the general finding the argument accepts. The publisher objects that it does not apply to them. The second bold is evidence that the objection fails. That is A. D and E both cast the portions as supporting the publisher, which reverses the argument's direction entirely.
Which of the following, if true, most seriously weakens the commentator's conclusion?
Reverse causation: if better memory drives more reading, the entire correlation is accounted for with no effect of reading on memory. That is the strongest of the four alternatives from V2.8 available here.
A raises a measurement concern — real, but it does not explain the association away. C and E are unmeasured variables with no stated direction. D is true of the population generally and does not distinguish the two groups.
Management's conclusion depends on which of the following assumptions?
The premise is revenue per occupied room. The conclusion is about revenue overall. Those come apart the moment occupancy changes: constant per-room revenue with fewer rooms occupied means total revenue fell.
Negate E — occupancy did decline — and the conclusion fails while the premise stays true. That is the definition of a necessary assumption. A is about preference, B about product mix, C about cost structure, D about the wider market; negate any of them and the revenue reasoning survives intact.
The answer to which of the following questions would be most useful in evaluating the director's conclusion?
Two-way test. If the eliminated hours were quiet ones, then a 3 percent loan drop is exactly what you would expect and the director is right. If they were busy hours, then a 3 percent drop across a 15 percent cut means something else is going on — displaced demand, or loans concentrated elsewhere — and the conclusion is unsafe.
B is about the financial target, not about service. C is about staff. D is worth knowing but does not distinguish the two scenarios: whichever way it is answered, the question of which hours were cut still decides the argument. E is an external comparison.
All eight types are questions about the same assumption. Find the gap first, in your own words; the stem only tells you what to do with it.
V3 moves to Reading Comprehension, and starts where this chapter started — with structure rather than content. The map you build there is what V4's question types depend on.