01 / Probability and risk ideasEvery board · Both tiers
An event with probability 0 is:
- Aimpossible
- Bcertain
- Cunlikely
- Dvery unlikely
Show the answer
Answer: A impossible
Probabilities run from 0 (impossible) to 1 (certain), or 0% to 100%. A value of 0.5 is an even chance. A probability above 1 means there is a mistake.
Checked against assets.publishing.service.gov.uk
02 / Probability and risk ideasEvery board · Both tiers
If P(A) is the probability of A, the probability that A does not happen is:
- A1 plus P(A)
- B1 minus P(A)
- CP(A) minus 1
- DP(A) divided by 2
Show the answer
Answer: B 1 minus P(A)
A either happens or it does not, so the two probabilities add up to 1. That makes the probability of “not A” equal to 1 − P(A).
Checked against assets.publishing.service.gov.uk
03 / Probability and risk ideasEvery board · Both tiers
As the number of trials increases, the relative frequency of an event usually:
- Amoves further from the true probability
- Bstays exactly the same
- Cgets closer to the true probability
- Dbecomes exactly 0.5
Show the answer
Answer: C gets closer to the true probability
Relative frequency is the fraction of trials in which the event happened. With more trials it settles closer to the true probability, so a large number of trials gives the best estimate.
Checked against assets.publishing.service.gov.uk
04 / Probability and risk ideasEvery board · Both tiers
Theoretical probability can be used when:
- Athere is no information at all
- Bthe outcomes are continuous
- Cthe dice is known to be biased
- Dall the outcomes are equally likely
Show the answer
Answer: D all the outcomes are equally likely
When outcomes are equally likely, such as on a fair dice, probability can be worked out in theory. When they are not, such as a drawing pin, an experiment gives an estimate.
Checked against assets.publishing.service.gov.uk
05 / Probability and risk ideasEvery board · Both tiers
A dice is rolled 600 times. It lands on six 250 times.
What does this suggest?
- AThe dice may be biased
- BA six is impossible
- CMore rolls would make it fair
- DThe dice is fair
Show the answer
Answer: A The dice may be biased
For a fair dice you would expect about 100 sixes in 600 rolls. Getting 250 is far more than expected, which suggests the dice is biased. Comparing results with theory is how bias is spotted.
Checked against assets.publishing.service.gov.uk
06 / Probability and risk ideasEvery board · Higher only
In 200 spins, a spinner lands on red slightly less often than theory predicts.
What is the fairest conclusion?
- AThe spinner is definitely biased against red
- BThe theory must be wrong for this spinner
- CRed is impossible
- DThe gap could be chance, so more trials would help
Show the answer
Answer: D The gap could be chance, so more trials would help
Experimental results rarely match theory exactly. A small gap may just be chance; a large gap over many trials is stronger evidence of bias. Say “may be biased” rather than “definitely”.
Checked against assets.publishing.service.gov.uk
07 / Probability and risk ideasEvery board · Both tiers
How do you find the number of times an outcome is expected to happen?
- AAdd its probability to the total number of trials
- BDivide its probability by the number of trials
- CMultiply its probability by the number of trials
- DMultiply the number of trials by 2
Show the answer
Answer: C Multiply its probability by the number of trials
Expected frequency = probability × number of trials (or size of the group). It is what you would expect on average, not a promise of exactly that number.
Checked against assets.publishing.service.gov.uk
08 / Probability and risk ideasEvery board · Both tiers
The absolute risk of an outcome for a group is:
- Aone group’s risk divided by another’s
- Bthe probability it happens to someone in that group
- Cthe chance it never happens
- Dthe number of people in the group
Show the answer
Answer: B the probability it happens to someone in that group
Absolute risk is the probability for one group. Relative risk divides one group’s risk by another’s, in the order the question asks, to show how many times as likely the outcome is.
Checked against assets.publishing.service.gov.uk
09 / Probability and risk ideasEvery board · Both tiers
Smokers have a relative risk of 2 for a disease compared with non-smokers. What does this mean?
- ASmokers are 200% more likely to get it
- BSmokers will definitely get it
- CTwo smokers in every 100 get it
- DSmokers are twice as likely to get it
Show the answer
Answer: D Smokers are twice as likely to get it
A relative risk of 2 means twice as likely, which is 100% more likely, not 200%. A relative risk of 1 means no difference between the groups.
Checked against assets.publishing.service.gov.uk
10 / Probability and risk ideasEvery board · Both tiers
A news story says a habit “doubles the risk” of a rare illness, from 1 in 10,000 people to 2 in 10,000.
Why is it useful to know the absolute risk here?
- AIt shows the relative risk is wrong
- BIt shows the habit is very dangerous
- CThe illness stays rare even though the risk doubles
- DIt proves the habit causes the illness
Show the answer
Answer: C The illness stays rare even though the risk doubles
A relative risk can sound alarming on its own. The absolute risks show the real chance is still tiny: 2 in 10,000. Good reporting gives both.
Checked against assets.publishing.service.gov.uk
11 / Probability and risk ideasEvery board · Both tiers
Two events are mutually exclusive if:
- Aone does not affect the chance of the other
- Bthey always happen together
- Cthey cannot happen at the same time
- Dthey have the same probability
Show the answer
Answer: C they cannot happen at the same time
Mutually exclusive events cannot both happen at once: 2 is not odd, so one roll cannot be both. Do not mix this up with independence, which is about one event not affecting the other.
Checked against assets.publishing.service.gov.uk
12 / Probability and risk ideasEvery board · Both tiers
Two events are independent if:
- Atheir probabilities always add up to exactly 1
- Bone happening does not change the chance of the other
- Cthey always happen together
- Dthey cannot both happen at the same time
Show the answer
Answer: B one happening does not change the chance of the other
Independent events do not affect each other: a coin cannot change how a dice lands. Picks without replacement are not independent, because the first pick changes what is left.
Checked against assets.publishing.service.gov.uk
13 / Probability and risk ideasEvery board · Both tiers
For independent events A and B, P(A and B) equals:
- AP(A) divided by P(B)
- BP(A) plus P(B)
- CP(A) times P(B)
- DP(A) minus P(B)
Show the answer
Answer: C P(A) times P(B)
For independent events you multiply the probabilities. Knowing B has happened does not change the chance of A, so P(A|B) = P(A). Either fact can be used to test for independence.
Checked against assets.publishing.service.gov.uk
14 / Probability and risk ideasEvery board · Both tiers
For mutually exclusive events A and B, P(A or B) equals:
- AP(A) divided by P(B)
- BP(A) plus P(B)
- CP(A) times P(B)
- DP(A) minus P(B)
Show the answer
Answer: B P(A) plus P(B)
Mutually exclusive events cannot happen together, so P(A and B) = 0 and you can add the probabilities to find P(A or B).
Checked against assets.publishing.service.gov.uk
15 / Probability and risk ideasEvery board · Higher only
Why is P(A and B) subtracted in P(A or B) = P(A) + P(B) − P(A and B)?
- ASo the overlap is not counted twice
- BBecause A and B must be independent
- CSo the answer is always 1
- DBecause the events are exhaustive
Show the answer
Answer: A So the overlap is not counted twice
When A and B can happen together, adding P(A) and P(B) counts the overlap twice, as a Venn diagram shows. Subtracting P(A and B) once corrects this.
Checked against qualifications.pearson.com
16 / Probability and risk ideasEvery board · Both tiers
A set of events is exhaustive if:
- Abetween them they cover every possible outcome
- Bnone of them can happen
- Cthey can all happen at once
- Dthey all have the same probability
Show the answer
Answer: A between them they cover every possible outcome
Exhaustive events cover everything that can happen. If they are also mutually exclusive, exactly one of them must happen, so their probabilities add up to 1.
Checked against assets.publishing.service.gov.uk
17 / Probability and risk ideasEvery board · Both tiers
P(B|A) means:
- Athe probability of A or B
- Bthe probability of both A and B
- Cthe probability of A, given that B has happened
- Dthe probability of B, given that A has happened
Show the answer
Answer: D the probability of B, given that A has happened
The bar is read as “given”. P(B|A) is a conditional probability: the chance of B when we already know A has happened. In GCSE Statistics this is on both tiers.
Checked against assets.publishing.service.gov.uk
18 / Probability and risk ideasEvery board · Both tiers
Two sweets are taken from a bag, one after the other, without the first being put back. Why are the picks not independent?
- AThe first pick changes what is left in the bag
- BOnly two sweets are taken out of the bag
- CThe bag is chosen at random
- DSweets come in different colours
Show the answer
Answer: A The first pick changes what is left in the bag
Without replacement, the first sweet taken changes the contents of the bag, so the probabilities for the second pick depend on the first. That is why tree diagrams change their fractions on the second branch.
Checked against assets.publishing.service.gov.uk
19 / Probability and risk ideasEvery board · Higher only
Which is a condition for a binomial model?
- AA fixed number of trials
- BA continuous outcome
- CA probability that changes each trial
- DTrials that depend on each other
Show the answer
Answer: A A fixed number of trials
A binomial model needs a fixed number of trials, two outcomes each time (success or failure), the same probability every trial, and independent trials. State them in context.
Checked against assets.publishing.service.gov.uk
20 / Probability and risk ideasEvery board · Higher only
A pupil rolls a dice until she gets a six, and counts the rolls.
Why is a binomial model not suitable?
- AThe chance of a six changes
- BThe rolls are not independent
- CThe dice has more than two outcomes
- DThere is no fixed number of trials
Show the answer
Answer: D There is no fixed number of trials
The number of rolls is not fixed in advance: she stops when the six comes. A binomial model needs a fixed number of trials. Each roll can still be “six” or “not six”.
Checked against qualifications.pearson.com
21 / Probability and risk ideasEvery board · Higher only
Five cards are taken from a pack of 52 without replacement, and the number of hearts is counted.
Why is a binomial model not suitable?
- AThere are more than two suits
- BCards are qualitative data
- CThe chance of a heart changes after each card
- DThere is no fixed number of trials
Show the answer
Answer: C The chance of a heart changes after each card
Without replacement, each card taken changes the chance that the next is a heart, so the probability is not constant and the trials are not independent.
Checked against assets.publishing.service.gov.uk
22 / Probability and risk ideasEvery board · Higher only
Each packet of cards has the same chance of holding a rare card, whatever is in other packets. Ali buys 5 packets.
Which feature of this situation fits a binomial model?
- AEach packet either has a rare card or not
- BThe packets cost different amounts
- CRare cards are worth more money
- DAli chooses which packets to buy
Show the answer
Answer: A Each packet either has a rare card or not
Here there are 5 trials (packets), two outcomes each (rare card or not), the same chance each time, and packets do not affect each other. Examiners want the conditions given in the context.
Checked against cdn.sanity.io
23 / Probability and risk ideasEdexcel · Higher only
In the notation B(n, p), n stands for:
- Athe probability of success
- Bthe number of trials
- Cthe number of successes
- Dthe mean
Show the answer
Answer: B the number of trials
B(n, p) is a binomial distribution with n trials and probability p of success on each. Its mean is np: for example, 10 trials with p = 0.3 gives a mean of 3 successes.
Checked against qualifications.pearson.com
24 / Probability and risk ideasEvery board · Higher only
The normal distribution is:
- Aflat and rectangular
- Bpositively skewed
- Csymmetrical and bell-shaped
- Dnegatively skewed
Show the answer
Answer: C symmetrical and bell-shaped
A normal distribution models continuous data with a symmetrical bell shape. Because it is symmetrical, the mean, median and mode are all at the centre.
Checked against assets.publishing.service.gov.uk
25 / Probability and risk ideasEvery board · Higher only
For a normal distribution, about what percentage of values lie within 1 standard deviation of the mean?
- A50%
- B68%
- C95%
- D99.7%
Show the answer
Answer: B 68%
For normal data, about 68% of values lie within 1 standard deviation of the mean, about 95% within 2, and almost all (99.7%) within 3. These facts are not given in the exam.
Checked against assets.publishing.service.gov.uk
26 / Probability and risk ideasEvery board · Higher only
For normally distributed data, a value more than 3 standard deviations from the mean is:
- Aimpossible
- Bvery unusual
- Cquite common
- Dthe most likely value
Show the answer
Answer: B very unusual
Only about 0.3% of normal data lies more than 3 standard deviations from the mean. Such a value is very unusual, so it is worth checking as a possible outlier, but it is not impossible.
Checked against assets.publishing.service.gov.uk
27 / Probability and risk ideasEvery board · Higher only
Two normal curves have their peaks at the same place on the horizontal axis. Curve A is tall and narrow; curve B is low and wide.
What does this show?
- ASame mean, but B has the larger standard deviation
- BA has the larger mean
- CB has the larger mean
- DSame mean, but A has the larger standard deviation
Show the answer
Answer: A Same mean, but B has the larger standard deviation
The centre of a normal curve, its line of symmetry, is the mean. A wider, flatter curve has a larger standard deviation. The height of the peak is not the mean.
Checked against qualifications.pearson.com
28 / Probability and risk ideasEvery board · Higher only
Which of these is most likely to follow a normal distribution?
- AThe number of cars each family owns
- BThe incomes of UK households
- CThe heights of adult women in the UK
- DThe scores from rolling one dice
Show the answer
Answer: C The heights of adult women in the UK
Heights of adults are continuous and roughly symmetrical, so they fit a normal model. Incomes are positively skewed, car counts are discrete, and dice scores are evenly spread.
Checked against cdn.sanity.io
29 / Probability and risk ideasEdexcel · Higher only
In the notation N(μ, σ²), the second number is:
- Athe standard deviation
- Bthe variance
- Cthe mean
- Dthe median
Show the answer
Answer: B the variance
N(μ, σ²) is a normal distribution with mean μ and variance σ². The standard deviation is the square root of the variance, so N(50, 16) has a standard deviation of 4.
Checked against qualifications.pearson.com
30 / Probability and risk ideasEvery board · Higher only
A standardised score tells you:
- Ahow many standard deviations a value is from the mean
- Bhow many marks a value is above or below the median
- Cthe percentage of marks a pupil got
- Dthe rank of a value when the data is in order
Show the answer
Answer: A how many standard deviations a value is from the mean
Standardised score = (value − mean) ÷ standard deviation. It puts results from different tests on the same scale, so they can be compared fairly.
Checked against qualifications.pearson.com
31 / Probability and risk ideasEvery board · Higher only
Sam’s standardised score in English is 1.5 and in maths is 0.8.
In which subject did Sam do better compared with the rest of the group?
- AIt cannot be compared
- BEnglish
- CEqually well in both
- DMaths
Show the answer
Answer: B English
Sam is 1.5 standard deviations above the mean in English but only 0.8 above in maths. The higher standardised score shows the better performance relative to the group.
Checked against assets.publishing.service.gov.uk
32 / Probability and risk ideasEvery board · Higher only
In a race a lower time is better. Ana’s standardised score in the 100 m is −1.2 and in the 200 m is −0.4.
In which race did Ana do better compared with the other runners?
- AThe 100 m
- BThe 200 m
- CIt cannot be compared
- DEqually well in both
Show the answer
Answer: A The 100 m
A negative standardised time means faster than the mean. −1.2 is further below the mean than −0.4, so Ana did better in the 100 m. With times, more negative is better.
Checked against assets.publishing.service.gov.uk
33 / Probability and risk ideasEvery board · Higher only
Compared with single values from a population, the means of samples from it are:
- Amore spread out
- Bspread out by the same amount
- Cless spread out
- Dalways equal to each other
Show the answer
Answer: C less spread out
In a sample, high and low values tend to balance out, so sample means vary less than single values. That is why quality control checks the means of small samples.
Checked against assets.publishing.service.gov.uk
34 / Probability and risk ideasEvery board · Higher only
Why does a factory take small samples from a production line at regular times?
- ATo test every item made
- BTo find the mode
- CTo check the process stays on target
- DTo make the items heavier
Show the answer
Answer: C To check the process stays on target
Quality control means checking samples regularly and plotting a summary, such as the mean, on a control chart. The middle line is the target; warning and action lines sit above and below it.
Checked against assets.publishing.service.gov.uk
35 / Probability and risk ideasEvery board · Higher only
A sample mean is plotted inside the warning lines. What should happen?
- ATake another sample straight away
- BCarry on as normal
- CMove the target line
- DStop the process
Show the answer
Answer: B Carry on as normal
Inside the warning lines: carry on. Between a warning and an action line: take another sample at once. Beyond an action line: stop and adjust or reset the machine.
Checked against cdn.sanity.io
36 / Probability and risk ideasEdexcel · Higher only
On a control chart, the warning lines are usually drawn:
- Aat the highest and lowest values
- B1 standard deviation from the target
- C2 standard deviations from the target
- D3 standard deviations from the target
Show the answer
Answer: C 2 standard deviations from the target
Warning lines go 2 standard deviations (of the sample means) either side of the target, and action lines 3 standard deviations. Results beyond these are unlikely if the process is working properly.
Checked against qualifications.pearson.com
37 / Probability and risk ideasEdexcel · Higher only
A control chart of sample ranges is used to check:
- Awhether the process has become too variable
- Bthe cost of each item
- Cwhether the mean is on target
- Dhow many items the machine makes each hour
Show the answer
Answer: A whether the process has become too variable
Charts can be drawn for sample means, medians or ranges. A range chart watches the spread: a large range means the items are too uneven, even if their mean is on target.
Checked against qualifications.pearson.com
38 / Probability and risk ideasEvery board · Both tiers
An insurance company uses past claims to estimate how many claims to expect next year. What is it using?
- AA natural experiment
- BRelative frequency to predict outcomes
- CA weighted index
- DA census of every customer
Show the answer
Answer: B Relative frequency to predict outcomes
Past data gives a relative frequency for each kind of claim, which estimates the probability. Multiplying by the number of customers gives the expected number of claims, which helps set prices.
Checked against assets.publishing.service.gov.uk