BioMedical Admissions Test (BMAT) Section 1: Thinking Skills Syllabus

Define the 'sweeping generalisation' (or hasty generalisation) flaw.

Hasty generalisation draws a broad conclusion from too small or unrepresentative a sample. Sweeping generalisation misapplies a general rule to an exceptional case where it should not hold. Both wrongly move between the general and the specific.

Define the 'straw man' flaw.

Misrepresenting or oversimplifying an opponent's position into a weaker version, then refuting that distorted version. It gives the illusion of defeating the argument while leaving the actual position untouched.

Define the flaw of 'confusing necessary and sufficient conditions'.

Treating a condition that is merely necessary (required) as if it were sufficient (enough on its own), or vice versa. E.g. oxygen is necessary for fire but not sufficient; assuming oxygen guarantees fire is this flaw.

What does it mean to 'strengthen' an argument, and what kind of statement achieves it?

To strengthen is to add a statement that makes the conclusion more likely to follow from the premises - typically by supporting an assumption, supplying confirming evidence, or ruling out an alternative explanation. It need not prove the conclusion, only increase support.

What does it mean to 'weaken' an argument?

To weaken is to add a statement that makes the conclusion less likely to follow - typically by undermining an assumption, introducing a counterexample, providing an alternative explanation, or casting doubt on the evidence. It need not disprove the conclusion.

In a strengthen/weaken question, why must you first identify the conclusion and the key assumption?

Because strengthening or weakening operates on the gap between premises and conclusion. Without pinpointing what the argument claims and what unstated assumption it relies on, you cannot tell which option supports or attacks that specific link.

What distinguishes a 'valid conclusion that can be drawn' from a merely plausible statement in BMAT?

A conclusion that can be 'safely drawn' or 'reliably concluded' must be fully supported by the given information with no extra assumptions - it follows necessarily. A plausible but unsupported statement, even if likely true in real life, is not a valid drawn conclusion.

What is the 'too strong' trap when selecting a conclusion that can be drawn?

An option overstates what the evidence supports - using absolute words like 'all', 'always', 'never', 'must' or 'proves' when the passage only supports a qualified claim ('some', 'may', 'suggests'). Such options go beyond the evidence and are invalid.

When interpreting a table, what should you check first before reading any value?

Read the title, the row and column headings, and especially the units and any multiplier (e.g. 'figures in thousands' or '%'). Misreading units or scale is the single most common error in data-interpretation questions.

On a graph, what is the danger of not checking whether an axis starts at zero?

A non-zero (truncated) baseline exaggerates differences between bars or the steepness of a line, making small changes look large. Always read the axis start value before judging the magnitude of any difference or trend.

How do you calculate a percentage change between an old value and a new value?

$$\text{percentage change} = \frac{\text{new} - \text{old}}{\text{old}} \times 100\%$$ A positive result is an increase, a negative result a decrease.

In a pie chart, how is the angle of a sector related to the proportion it represents?

The sector angle is the proportion multiplied by the full circle: $$\text{angle} = \frac{\text{category value}}{\text{total}} \times 360^{\circ}$$ Conversely, proportion $= \frac{\text{angle}}{360^{\circ}}$.

When reasoning from quantitative evidence, what is the difference between absolute and relative (percentage) figures, and why does it matter?

An absolute figure is a raw count; a relative figure expresses it as a fraction or percentage of a total. A large absolute number can be a small percentage and vice versa, so claims must be checked against the right measure - mixing them is a classic data flaw.

How should you evaluate a claim of the form 'data show X causes Y'?

Check whether the data actually show causation or only association/correlation. Ask: is the sample representative, is there a control or comparison group, could a third variable explain the link, and could the direction of causation be reversed? Most data only support correlation.

What is the base-rate consideration when evaluating a claim like 'most accident victims wore seatbelts, so seatbelts are dangerous'?

You must compare against the base rate - the proportion of all drivers who wear seatbelts. If almost everyone wears them, most victims will too even if belts reduce risk. Raw counts without the relevant denominator give a misleading conclusion.

State the basic probability of an event for equally likely outcomes.

$$P(\text{event}) = \frac{\text{number of favourable outcomes}}{\text{total number of equally likely outcomes}}$$ This value always satisfies $0 \leq P \leq 1$.

What is the addition rule for the probability of A or B when the events are mutually exclusive?

For mutually exclusive events (they cannot both occur): $$P(A \text{ or } B) = P(A) + P(B)$$ More generally, $P(A \text{ or } B) = P(A) + P(B) - P(A \text{ and } B)$.

What is the multiplication rule for the probability of two independent events both occurring?

For independent events A and B: $$P(A \text{ and } B) = P(A) \times P(B)$$ Independence means the occurrence of one does not change the probability of the other.

How do you find the probability of the complement of an event A (i.e. A not occurring)?

$$P(\text{not } A) = 1 - P(A)$$ This 'complement trick' is often the fastest route when calculating 'at least one' probabilities.

What is the fastest way to compute the probability of 'at least one' success in repeated independent trials?

Use the complement: $$P(\text{at least one}) = 1 - P(\text{none})$$ where $P(\text{none})$ is the product of the failure probabilities across all trials. This avoids summing many separate cases.

Distinguish 'probability', 'risk' and 'odds' as they appear in BMAT quantitative arguments.

Probability/risk is favourable outcomes over total outcomes, $\frac{a}{a+b}$ (a value $0$ to $1$). Odds is favourable to unfavourable, $\frac{a}{b}$ (e.g. 3:1). They are not interchangeable: odds of 3:1 corresponds to a probability of $\frac{3}{4}$, not $\frac{3}{1}$.

In Section 1, why must you treat the answer options themselves as a source of information?

Because BMAT is multiple-choice, options constrain the answer: you can back-substitute them, eliminate impossible ones, and infer the required precision. Using the options actively often turns a hard calculation into a quick test of candidates.

What is the recommended time-per-question pace for BMAT Section 1?

Section 1 historically allowed 35 questions in 60 minutes, giving roughly $\frac{60}{35} \approx 1.7$ minutes (about 100 seconds) per question. The practical rule is to average under two minutes per question to leave time for harder ones.

What is 'question triage' and why is it essential when every question carries equal marks?

Triage is rapidly sorting questions into do-now (quick, confident), do-later (solvable but slow), and guess-and-flag (very hard). Since every question scores one mark with no negative marking, securing all the easy marks first maximises score for the time available.

Why should you never leave any BMAT multiple-choice question blank?

Because there is no negative marking - wrong answers are not penalised. Any blank scores zero, whereas a guess has a positive expected value (e.g. $\frac{1}{4}$ chance on four options), so every question should at least be guessed.

With four options and no penalty, what is the expected mark from a pure random guess, and how does elimination change it?

A random guess yields $\frac{1}{4} = 0.25$ marks expected. Eliminating one wrong option raises the chance to $\frac{1}{3} \approx 0.33$; eliminating two raises it to $\frac{1}{2} = 0.5$. Elimination directly increases the expected value of a guess.

State the 'elimination technique' for narrowing multiple-choice options.

Discard options that are impossible or contradict the data (wrong sign, wrong order of magnitude, ruled out by a constraint), then choose among survivors. Even partial elimination improves guessing odds, and often only one option survives all constraints.

What is a sound rule for managing time when you get stuck on a single hard question?

Set a personal cutoff (e.g. about two minutes); if you have no clear route by then, mark your best-guess answer, flag the question, and move on. Returning later with fresh eyes or spare time is more efficient than burning minutes on one mark.

When a data-interpretation question gives more information than needed, what is the disciplined approach?

First identify the single quantity the question asks for, then locate only the figures that feed that quantity, ignoring all other rows, columns or graph series. Selecting relevant information before calculating prevents both wasted time and errors from distractor data.

Why is it important to read the exact wording of what a Critical Thinking question asks (e.g. 'best supports' vs 'is an assumption' vs 'is a flaw')?

Each instruction targets a different logical relationship - supporting evidence, an unstated necessary premise, or a reasoning error. The correct option for one instruction is wrong for another, so misreading the task type is a guaranteed loss of the mark despite correct understanding of the passage.