IMAT Logical Reasoning · Critical Thinking
Identifying Flaws in Reasoning
A flaw is a mistake in the step from reasons to conclusion; IMAT asks you to name it in plain words.
Why this matters
Flaw questions were asked 19 times between 2011 and 2022. The right answer was most often a causal mistake (a link read as a cause, the cause running the other way, another explanation ignored), and next a conclusion drawn from one example or an unrepresentative statistic. Attacking a person's motive and misreading the other side's argument also appear.
Concept 1 of 4: Causal flaws: correlation, reverse cause and ignored explanations
When two things go together, it is tempting to say one causes the other. But a third thing may cause both, the cause may run the other way, or the timing may be coincidence. This jump from 'goes with' to 'causes' is the flaw IMAT tests most often.
Definition
Look for a conclusion with causes, leads to, makes, is responsible for, because of, then check what the evidence really shows.
- Correlation is not causation: A and B happening together does not show that A causes B.
- Reverse causation: B may cause A (success may build confidence, not the other way).
- Third factor: something else, C, may cause both A and B.
- Post hoc: B happened after A, so A must have caused B.
- Ignored alternative: another explanation for the evidence is not considered.
- IMAT words these as 'It confuses a correlation with a cause', 'It assumes that X is responsible for Y, and not the other way round', 'It ignores other reasons for ...'.
Worked example
Name the flaw. 'A survey found that people who own a dog visit their doctor less often than people who do not. Owning a dog therefore makes people healthier, and doctors should encourage their patients to get one.'
Practice this conceptself-check · 3 quick reps
A link is only a flaw when the conclusion claims a cause
If the passage says only that two things are linked, or that one may contribute to the other, it has not confused correlation with causation. Check the wording of the conclusion before choosing that option.
Concept 2 of 4: Generalising from a small or unrepresentative sample
A conclusion about a whole group needs evidence from a fair slice of that group. One striking case, a handful of people, or people who chose to answer a survey can all give a false picture, however large the numbers look.
Definition
Ask two questions: who was studied, and who is the conclusion about? If they differ, there is a flaw.
- Hasty generalisation: a claim about all from one case or a few.
- Unrepresentative sample: the people studied differ from the group in the conclusion (one city, one age group, members of a club, readers of one website).
- Self-selection: people who choose to reply to a survey often have strong views.
- Anecdote against statistics: one personal story set against evidence about many people.
- A large sample does not fix a biased one.
Worked example
Name the flaw. 'A cooking website asked its readers whether they would pay more for organic vegetables. Of the 4,000 readers who replied, 3,000 said yes. So three quarters of the country's shoppers would pay more for organic food.'
Practice this conceptself-check · 3 quick reps
A big sample can still be a biased sample
Thousands of replies sound convincing, but if the people who replied differ from the group in the conclusion, the result does not transfer. IMAT options that praise or attack the sample's size miss the real point: who was in it.
Concept 3 of 4: Confusing necessary and sufficient conditions
A necessary condition must be met for something to happen; a sufficient condition is enough on its own to make it happen. Having a ticket is necessary to see a concert, but it is not sufficient: the concert must also take place. Arguments that mix the two sound right and do not follow.
Definition
Necessary: B is necessary for A if A cannot happen without B. Signals: 'only if B', 'A requires B', 'without B, no A'.
- Sufficient: B is sufficient for A if B guarantees A. Signals: 'if B, then A', 'B is enough for A'.
- 'If A then B' makes A sufficient for B and B necessary for A.
- Valid: if A then B; A; so B. And: if A then B; not B; so not A.
- Invalid (affirming the consequent): if A then B; B; so A.
- Invalid (denying the antecedent): if A then B; not A; so not B.
- Meeting a necessary condition never guarantees the result.
Worked example
Find the flaw. 'To board an international flight you must show a valid passport. Ravi has a valid passport, so he will be allowed to board his flight to Lisbon tomorrow.' Then compare: 'Mia was allowed to board her international flight, so she must have shown a valid passport.'
Practice this conceptself-check · 3 quick reps
'Only if' introduces the necessary condition
'You may enter only if you have a ticket' means a ticket is necessary, not that having one lets you in. Students often read 'only if' as 'if' and then accept a conclusion that does not follow.
Concept 4 of 4: Other named flaws: attacking the person, straw man, false choice and more
These flaws do not misread evidence; they replace reasoning with something else. They attack the speaker, distort the other side's view, pretend there are only two options, or quietly assume what they set out to prove. Learn each name together with the plain description IMAT would use in an option.
Definition
Each flaw, with how IMAT would describe it:
- Ad hominem: attacks the person's motive, character or habits instead of their reasons. 'It attacks the speaker rather than her argument.'
- Straw man: misrepresents the opposing view as weaker or more extreme, then attacks that version.
- False dichotomy (false choice): presents two options as the only ones when there are others.
- Slippery slope: claims one step will lead, through a chain of unsupported steps, to an extreme result.
- Circular reasoning: the conclusion is already assumed in the premises.
- Appeal to authority: true because an expert said so, especially an expert in another field.
- Appeal to popularity: true or right because many people believe or do it.
Worked example
Name the two flaws. 'Dr Ferri argues that energy drinks should not be sold to under-16s because of their high caffeine content. But Dr Ferri herself drinks several cups of strong coffee a day, so we need not take her argument seriously. Besides, if we ban energy drinks for teenagers, the next step will be banning coffee, chocolate and cola for everyone.'
Practice this conceptself-check · 3 quick reps
Citing an expert is not automatically a flaw
Relying on a relevant expert is normal and reasonable. It becomes a flaw when the expert speaks outside their field, or when the claim rests on authority alone with no reason given. Do not choose 'appeal to authority' just because a scientist is mentioned.
Summary — formulas & gotchas at a glance
A revision cheat-sheet for the formulas and gotchas above. Click any concept name to jump back to its full explanation.
Watch out for (4)
- A link is only a flaw when the conclusion claims a cause→ Causal flaws: correlation, reverse cause and ignored explanations
- A big sample can still be a biased sample→ Generalising from a small or unrepresentative sample
- 'Only if' introduces the necessary condition→ Confusing necessary and sufficient conditions
- Citing an expert is not automatically a flaw→ Other named flaws: attacking the person, straw man, false choice and more
Drill the past-year questions on this topic
Open the bank with this topic pre-filtered.