The short answer
The reasoning flaws Section I loves to test — and how to name where an argument actually breaks.
Written and checked by GAMSAT tutors — not AI-generated.
Try the reasoning style
We treat forgetting as a failure — a lapse to be patched with reminders and records. Yet a mind that kept everything could not think; it would drown in the undifferentiated noise of every moment it had ever lived. To forget is not so much to lose information as to decide, mostly without our noticing, what was never worth keeping.
The author's argument relies most directly on which unstated assumption?
Pick an option to see how the tutor reasons to the answer — not just whether you were right.
Not quite — the answer is B.
Work backwards from the conclusion: a mind that ‘kept everything’ supposedly ‘could not think.’ That only follows if thinking means leaving most of experience out — so B is the premise the argument quietly rests on. A raises reliability, which the passage never weighs; C contradicts ‘mostly without our noticing’; D smuggles in a claim about intellect the passage never makes. The question rewards finding the hidden premise, not recalling a fact.
Section I rewards readers who can pinpoint where an argument's logic breaks. You don't need the Latin name — you need to see that the conclusion doesn't follow from the evidence. A handful of recurring flaws cover most of what you'll meet.
The big four to recognise
Correlation ≠ causation: two things move together, so one is claimed to cause the other. False dichotomy: only two options offered when more exist. Straw man: attacking a distorted version of the opponent's view. Ad hominem: attacking the person, not the argument.
The one they test most
Correlation vs causation is the GAMSAT's favourite. "Towns with more libraries have less crime, so libraries reduce crime" ignores a confounder (wealthier towns have both). Whenever a claim leaps from "linked" to "causes," check for a third factor or reversed direction.
Correlation vs causation
What the data shows
- Two variables change together
- A pattern or association
- Says nothing about WHY
What's being claimed
- One variable causes the other
- A mechanism / direction
- Needs a controlled comparison to support
Worked example
"My grandfather smoked and lived to 95, so the warnings about smoking are clearly overblown." What's the flaw?
Check yourself
"Either we cut the budget entirely or the company collapses." Which flaw is this?
Key takeaways
- Find where the conclusion outruns the evidence — that's the flaw.
- Correlation ≠ causation: look for confounders or reversed direction.
- False dichotomy: only two options offered when more exist.
- Straw man: the opposing view has been distorted before being attacked.
- Ad hominem: the person is attacked instead of their argument.
- Hasty generalisation: a sweeping claim from too small or biased a sample.
Practise this with real GAMSAT-style questions
Free account: a timed diagnostic, an AI tutor that explains every answer, essay marking on the official rubric, and a plan built around your weak spots.