UCAT Decision Making: Probabilistic reasoning, set 1
Judging a claim about chance from a short statistical passage, where each option pairs Yes or No with a reason and only one reason is right. Test standard. 10 questions, 10 marks, about 11 minutes.
Download the questions (PDF) Download with worked solutions (PDF)
- Answer every question. Questions are standalone.
- Multiple-statement questions are worth 2 marks; the real test gives 1 mark for a partly correct response. Other questions have exactly one correct answer.
- A basic calculator is allowed in this subtest.
- 11 mark
Priya commutes to Harrowfield General Hospital on two buses. The Route 7 bus, run by Valecross Coaches, arrives late on 1 in 5 mornings. The Route 12 bus, run by a different company, arrives late on 1 in 4 mornings. Whether one bus is late has no effect on whether the other is late.
Is the chance that both of Priya's buses are late on the same morning less than 10%?
Select the option that gives the correct answer and the correct reason.
- 21 mark
A packet of tomato seeds sold by Brindle Gardens states that each seed germinates with probability 0.8, independently of the others. Tomasz plants 3 seeds from the packet in one pot.
Is the chance that at least one of the three seeds germinates greater than 99%?
Select the option that gives the correct answer and the correct reason.
- 31 mark
A rack in the Northvale Trust laboratory holds 10 sample tubes, and 3 of them are contaminated. A technician picks 2 tubes at random, one after the other, without putting the first back.
Is the chance that both tubes picked are contaminated less than 1 in 12?
Select the option that gives the correct answer and the correct reason.
- 41 mark
Is a member of staff who was absent with a flu-like illness more likely to have been vaccinated than not vaccinated?
Select the option that gives the correct answer and the correct reason.
- 51 mark
A screening test used at the Ellerby Health Centre wrongly gives a positive result to 2% of healthy people who take it. This year 1,500 healthy people are expected to take the test.
Should the centre expect more than 25 false positive results from these 1,500 people this year?
Select the option that gives the correct answer and the correct reason.
- 61 mark
A health page reports that taking Supplement K 'halves the risk' of a rare eye condition. In the study it cites, 4 in every 10,000 people not taking the supplement developed the condition within five years, compared with 2 in every 10,000 people taking it.
Does taking Supplement K lower a person's five-year chance of developing the condition by 50 percentage points?
Select the option that gives the correct answer and the correct reason.
- 71 mark
Each night at Corrowby Infirmary a computer chooses one of 4 registrars at random to lead the night shift. Every registrar has the same chance each night, and each night's choice is made independently of every earlier night. Dr Osei has not been chosen on any of the last 6 nights.
Is Dr Osei more likely than 1 in 4 to be chosen tonight?
Select the option that gives the correct answer and the correct reason.
- 81 mark
Is a randomly chosen Year 13 student who applied to study medicine at least 3 times as likely to be taking Chemistry as not taking it?
Select the option that gives the correct answer and the correct reason.
- 91 mark
The small ferry to the island of Skellan is cancelled because of the weather on 1 in 10 days, and each day's weather is independent of every other day. Mei has booked a sailing on each of the next five days.
Is there more than a 50% chance that at least one of Mei's five sailings is cancelled?
Select the option that gives the correct answer and the correct reason.
- 101 mark
Last year the town of Harrowfield, with 40,000 residents, recorded 60 cycling injuries among its residents. The town of Bramley, with 10,000 residents, recorded 25.
Based on these figures, was a Harrowfield resident more likely to suffer a cycling injury last year than a Bramley resident?
Select the option that gives the correct answer and the correct reason.
Worked solutions
Each solution below shows the working. The official material for this test gives only the correct option letter.
Question 1
Answer or working Marks C 1 The passage says the two buses are independent: one being late has no effect on the other. For independent events, P(both) = P(first) x P(second) = 1/5 x 1/4 = 1/20. 1/20 = 5/100 = 5%, which is less than 10%, so the verdict is Yes. Option A also says Yes, but its reason (each is late on fewer than a third of mornings) would allow a joint chance above 10%, so it does not justify the claim. The answer is C. Why not A: Right verdict, wrong reason: two buses each late on under a third of mornings could still both be late on, say, 1/3 x 1/3 = 1/9 of mornings, which is over 10%. The reason does not settle the claim; only the product does. Why not B: Adds the two probabilities instead of multiplying them. 1/5 + 1/4 = 4/20 + 5/20 = 9/20 is roughly the chance that at least one is late (with the overlap counted twice), not the chance that both are. Why not D: Contradicts the passage, which says one bus being late has no effect on the other. Inventing a dependence the data rule out is not a valid reason. Question 2
Answer or working Marks A 1 'At least one germinates' is the complement of 'none germinates', so P(at least one) = 1 - P(none). Each seed fails with probability 1 - 0.8 = 0.2, independently, so P(none) = 0.2 x 0.2 x 0.2 = 0.008. P(at least one) = 1 - 0.008 = 0.992 = 99.2%, which is greater than 99%. So the verdict is Yes, with the complement as the reason. The answer is A. Why not B: 0.8^3 = 0.512 is the chance that ALL THREE seeds germinate. The claim is about AT LEAST ONE, which is a much more likely event. Why not C: Adds the probabilities: 0.8 + 0.8 + 0.8 = 2.4. A probability can never exceed 1, so adding is the wrong operation; the verdict is right only by accident. Why not D: Assumes the chance for a group cannot exceed the chance for one seed. Having three tries makes 'at least one success' more likely than any single success. Question 3
Answer or working Marks C 1 First pick: 3 of the 10 tubes are contaminated, so P(first contaminated) = 3/10. Without replacement, if the first was contaminated then 2 contaminated tubes remain among 9, so P(second contaminated | first contaminated) = 2/9. P(both) = 3/10 x 2/9 = 6/90 = 1/15. Since 15 > 12, 1/15 < 1/12 (about 0.067 against about 0.083). So the verdict is Yes, for the reason in option C. The answer is C. Why not A: Treats the picks as if the first tube were put back, so the second pick still has 3 contaminated tubes out of 10. Without replacement there are only 2 contaminated tubes left out of 9. Why not B: Gives the chance that ONE picked tube is contaminated, not that both are. Why not D: Right verdict, wrong reason: it reduces the contaminated count to 2 for the second pick but forgets the total also drops to 9, so 6/100 is not the correct probability. Question 4
Answer or working Marks B 1 The question conditions on being absent, so restrict attention to the 'Absent' column: 18 + 30 = 48 staff. Of those 48, 18 were vaccinated, so P(vaccinated | absent) = 18/48 = 3/8 = 37.5%. 37.5% is less than a half, so an absent member of staff is LESS likely to have been vaccinated than not; the verdict is No. Option C also says No, but 18/180 is the reverse conditional (the chance of absence given vaccination), which does not answer this question. The answer is B. Why not A: Uses the share of ALL staff who were vaccinated. The question is about staff who were absent, so only the 'Absent' column counts. Why not C: Right verdict, wrong reason: 18/180 is P(absent | vaccinated), the reverse conditional. The share of vaccinated staff who were absent says nothing on its own about what share of absent staff were vaccinated; that needs the Absent column. Why not D: Reads the wrong column: 162/252 is the share vaccinated among staff who were NOT absent, not among those who were. Question 5
Answer or working Marks B 1 Expected frequency = probability x number of trials. 2% = 0.02, so the expected number of false positives is 0.02 x 1,500 = 30. 30 is more than 25, so the verdict is Yes, and option B gives the correct calculation. The answer is B. Why not A: Decimal slip: uses 0.002 instead of 0.02, giving 0.002 x 1,500 = 3. Why not C: Misunderstands expected frequency. Individual results are uncertain, but the expected number is still probability x number of trials. Why not D: Right verdict, wrong reason: uses 0.2 (20%) instead of 0.02, giving 0.2 x 1,500 = 300. Question 6
Answer or working Marks D 1 Convert both risks to percentages: 4/10,000 = 0.04% and 2/10,000 = 0.02%. The absolute fall is 0.04 - 0.02 = 0.02 percentage points. The relative fall is 0.02/0.04 = 50%, which is what 'halves the risk' means. A fall of 50 percentage points is impossible here, since the starting risk is only 0.04%. So the verdict is No. Option B also says No but gets the size of the fall wrong by a factor of 100. The answer is D. Why not A: Confuses a relative change with an absolute one. The risk does fall by 50% of itself, but a 50 percentage point fall would mean dropping from, for example, 50.04% to 0.04%. Why not B: Right verdict, wrong reason: 2 in 10,000 is 0.02%, not 2%, so the fall is 0.02 percentage points, not 2. Why not C: Misreads 'halves the risk' as 'protects half the people'. Only 4 in 10,000 would have developed the condition anyway, so at most 2 in 10,000 are spared. Question 7
Answer or working Marks A 1 The passage says each night's choice is independent of earlier nights and every registrar has the same chance. With 4 equally likely registrars, P(Dr Osei tonight) = 1/4, whatever happened on previous nights. So he is not more likely than 1 in 4 to be chosen; the verdict is No, because of independence. Option C also says No but relies on bias the passage rules out. The answer is A. Why not B: The gambler's fallacy: a random process with no memory does not 'owe' anyone a turn. Earlier nights do not change tonight's probability. Why not C: Right verdict, wrong reason: the passage states every registrar has the same chance each night, so the draw is not biased. Six misses in a row is unremarkable, with chance (3/4)^6 = 729/4096, about 18%. Why not D: (3/4)^7 = 2187/16384 is the chance, judged BEFORE the run began, of seven misses in a row. Six misses have already happened, so tonight's chance of a miss is simply 3/4 and of a pick 1/4. Question 8
Answer or working Marks D 1 The question conditions on having applied to medicine, so use only the 'Applied' column: 24 + 6 = 30 students. Of the 30 applicants, 24 take Chemistry and 6 do not, so taking Chemistry is 24/6 = 4 times as likely as not (P = 24/30 = 4/5). 4 times is at least 3 times, so the verdict is Yes. Option C also says Yes but reasons from the non-applicants. The answer is D. Why not A: Uses the reverse conditional: 24/80 is P(applied | takes Chemistry), not P(takes Chemistry | applied). Why not B: Uses the overall application rate. The question is only about students who applied, so the 170 who did not are irrelevant. Why not C: Right verdict, wrong reason: 114/120 describes students who did NOT apply, which is the wrong group. It says nothing directly about the applicants. Question 9
Answer or working Marks B 1 'At least one cancelled' is the complement of 'none cancelled', so P(at least one cancelled) = 1 - P(all five sail). Each day the sailing goes ahead with probability 0.9, independently, so P(all five sail) = 0.9^5 = 0.59049. P(at least one cancelled) = 1 - 0.59049 = 0.40951, about 41%, which is not more than 50%. The verdict is No. Option C also says No but reaches 0.5 by adding probabilities, which is invalid. The answer is B. Why not A: 0.9^5 = 0.59049 is the chance that NO sailing is cancelled. Reading it as the chance of at least one cancellation gets the event backwards. Why not C: Right verdict, wrong reason: adding 0.1 five times double-counts days with more than one cancellation. The correct figure is 0.40951, not 0.5; with ten sailings this method would give 1, which is plainly wrong. Why not D: 0.1^5 is the chance that ALL five sailings are cancelled, so 1 - 0.1^5 is the chance that at least one sails, not that at least one is cancelled. Question 10
Answer or working Marks A 1 To compare the chance for one resident, convert each count to a rate per head. Harrowfield: 60/40,000 = 1.5 per 1,000 (0.15%). Bramley: 25/10,000 = 2.5 per 1,000 (0.25%). A Bramley resident had the higher chance, so the verdict is No, even though Harrowfield had more injuries in total. Option C also says No but misstates the rates by a factor of ten. The answer is A. Why not B: Compares absolute counts. Harrowfield has four times as many residents, so the counts must be turned into rates before they can be compared. Why not C: Right verdict, wrong reason: 25/10,000 = 0.25% and 60/40,000 = 0.15%, not 2.5% and 1.5%. The percentages are ten times too large. Why not D: Adds an assumption the data do not contain. Nothing in the passage says anything about traffic, and the injury figures themselves point the other way.