Stratified Sampling (Higher) - Worksheets, Questions and Revision
56 original exam-style questions - 31 pages of questions with a full mark scheme - free printable PDF.
Stratified Sampling (Higher) revision guideMore GCSE Statistics practiceMock exams
This topic is chapter 1 of GCSE Statistics Higher Topic Workbook 1 and chapter 1 of GCSE Statistics Higher 10-Minute Tests 1.
1.5 Stratified Sampling (Higher)
Which calculation correctly gives the number of strings players to include in the sample?
- A) 96/240 x 30
- B) 30/96 x 240
- C) 96/30 x 240
- D) 240/96 x 30
| Group | Number |
|---|---|
| cows | 60 |
| sheep | 40 |
| pigs | 20 |
| goats | 80 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| teachers | 60 |
| technicians | 200 |
| administrators | 40 |
| cleaners | 200 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| cows | 120 |
| sheep | 30 |
| pigs | 150 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| students | 120 |
| adults | 80 |
| pensioners | 200 |
| children | 80 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| students | 100 |
| adults | 60 |
| pensioners | 20 |
| children | 20 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| juniors | 50 |
| seniors | 80 |
| veterans | 40 |
| guests | 30 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| juniors | 120 |
| seniors | 200 |
| veterans | 40 |
| guests | 40 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| Year 7 | 40 |
| Year 8 | 30 |
| Year 9 | 50 |
| Year 10 | 80 |
Work out how many should be sampled from each group.
(a) Work out the sampling fraction in its simplest form.
(b) Work out the total size of the sample.
(a) Work out the sampling fraction in its simplest form.
(b) Work out the total size of the sample.
| Group | Number |
|---|---|
| students | 60 |
| adults | 120 |
| pensioners | 60 |
| children | 60 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| students | 100 |
| adults | 200 |
| pensioners | 20 |
| children | 80 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| teachers | 20 |
| technicians | 50 |
| administrators | 80 |
Work out how many should be sampled from each group.
(a) Work out the proportion who cycle in each survey as a percentage, to 1 decimal place if needed.
(b) Which survey is likely to give the more reliable estimate, and why?
| Group | Number |
|---|---|
| juniors | 30 |
| seniors | 60 |
| veterans | 30 |
| guests | 30 |
Work out how many should be sampled from each group.
(a) Work out the sampling fraction in its simplest form.
(b) Work out the total size of the sample.
| Group | Number |
|---|---|
| teachers | 150 |
| technicians | 40 |
| administrators | 60 |
Work out how many should be sampled from each group.
(a) Work out the sampling fraction in its simplest form.
(b) Work out the total size of the sample.
| Group | Number |
|---|---|
| teachers | 50 |
| technicians | 30 |
| administrators | 20 |
| cleaners | 50 |
Work out how many should be sampled from each group.
(a) Work out the sampling fraction in its simplest form.
(b) Work out the total size of the sample.
| Group | Number |
|---|---|
| students | 150 |
| adults | 200 |
| pensioners | 150 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| students | 150 |
| adults | 20 |
| pensioners | 30 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| teachers | 30 |
| technicians | 100 |
| administrators | 120 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| teachers | 20 |
| technicians | 80 |
| administrators | 100 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| cows | 60 |
| sheep | 50 |
| pigs | 50 |
| goats | 40 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| juniors | 20 |
| seniors | 50 |
| veterans | 60 |
| guests | 120 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| Year 7 | 60 |
| Year 8 | 60 |
| Year 9 | 60 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| teachers | 50 |
| technicians | 60 |
| administrators | 40 |
| cleaners | 100 |
Work out how many should be sampled from each group.
| Group | Number |
|---|---|
| Year 7 | 40 |
| Year 8 | 40 |
| Year 9 | 120 |
Work out how many should be sampled from each group.
Which of these variables would be LEAST suitable to use when deciding the strata?
- A) Year group (Year 7 to Year 11)
- B) Gender
- C) The pupil's favourite colour
- D) Whether the pupil has school dinners or a packed lunch
| Age group | Under 10 | 10-13 | 14+ |
|---|---|---|---|
| Students | 72 | 63 | 45 |
A stratified sample of 20 students is needed for a feedback survey.
To choose them, the branch manager reads the following string of random digits in pairs:
40 91 07 40 66 12 84
| Ward | A&E | Maternity | Paediatrics | Surgery | Elderly Care |
|---|---|---|---|---|---|
| Nurses | 40 | 35 | 30 | 45 | 25 |
A stratified sample of 40 nurses is required, with each ward's sample size rounded to the nearest whole number.
| Study mode | Science | Humanities |
|---|---|---|
| Full-time | 38% | 22% |
| Part-time | 12% | 28% |
A stratified sample of 150 students is required, split across all four groups in proportion to their size.
The trainee's working is shown below.
Sampling fraction = 45/900 = 1/20
Speakers sample = 60 - 1/20 = 59.95, rounded to 60
Exhibitors sample = 240 x 1/20 = 12
Attendees sample = 600 x 1/20 = 30
Question 1
- A caoB1
- Answer: A) 96/240 x 30
Question 2
- (a) 156/260 oe (3/5, 0.6) caoB1
- (a) Answer: 156/260 (= 3/5)
- (b) 156/260 x 40, oeM1
- (b) 24 caoA1
- (b) Answer: 24
- (c) 104/260 x 40, oeM1
- (c) 16 caoA1
- (c) Answer: 16
Question 3
- sample size / population = 30/200M1
- group size x sampling fraction for at least one groupM1
- 9, 6, 3, 12 caoA1
- Answer: 9, 6, 3, 12
Question 4
- sample size / population = 100/500M1
- group size x sampling fraction for at least one groupM1
- 12, 40, 8, 40 caoA1
- Answer: 12, 40, 8, 40
Question 5
- sample size / population = 50/300M1
- group size x sampling fraction for at least one groupM1
- 20, 5, 25 caoA1
- Answer: 20, 5, 25
Question 6
- 5/80M1
- 1/16 x 1200M1
- 75 caoA1
- Answer: 75
Question 7
- sample size / population = 60/480M1
- group size x sampling fraction for at least one groupM1
- 15, 10, 25, 10 caoA1
- Answer: 15, 10, 25, 10
Question 8
- sample size / population = 40/200M1
- group size x sampling fraction for at least one groupM1
- 20, 12, 4, 4 caoA1
- Answer: 20, 12, 4, 4
Question 9
- 1000 x 61M1
- 61000 kg caoA1
- Answer: 61000 kg
Question 10
- sample size / population = 40/200M1
- group size x sampling fraction for at least one groupM1
- 10, 16, 8, 6 caoA1
- Answer: 10, 16, 8, 6
Question 11
- sample size / population = 40/400M1
- group size x sampling fraction for at least one groupM1
- 12, 20, 4, 4 caoA1
- Answer: 12, 20, 4, 4
Question 12
- 77/200M1
- 77/200 x 1000M1
- 385 caoA1
- Answer: 385
Question 13
- sample size / population = 60/200M1
- group size x sampling fraction for at least one groupM1
- 12, 9, 15, 24 caoA1
- Answer: 12, 9, 15, 24
Question 14
- 15/20M1
- 3/4 oeA1
- 80 x 3/4M1
- 60 caoA1
- Answer: (a) 3/4; (b) 60
Question 15
- 15/60M1
- 1/4 oeA1
- 400 x 1/4M1
- 100 caoA1
- Answer: (a) 1/4; (b) 100
Question 16
- sample size / population = 25/300M1
- group size x sampling fraction for at least one groupM1
- 5, 10, 5, 5 caoA1
- Answer: 5, 10, 5, 5
Question 17
- 3/100M1
- 3/100 x 3000M1
- 90 caoA1
- Answer: 90
Question 18
- sample size / population = 100/400M1
- group size x sampling fraction for at least one groupM1
- 25, 50, 5, 20 caoA1
- Answer: 25, 50, 5, 20
Question 19
- sample size / population = 30/150M1
- group size x sampling fraction for at least one groupM1
- 4, 10, 16 caoA1
- Answer: 4, 10, 16
Question 20
- 3/20 x 100 or 12/100 x 100M1
- 15% and 12% caoA1
- Survey BB1
- it has a larger sample (100 compared with 20)B1
- Answer: (a) 15% and 12%; (b) Survey B, larger sample
Question 21
- sample size / population = 100/150M1
- group size x sampling fraction for at least one groupM1
- 20, 40, 20, 20 caoA1
- Answer: 20, 40, 20, 20
Question 22
- 10/40M1
- 1/4 oeA1
- 160 x 1/4M1
- 40 caoA1
- Answer: (a) 1/4; (b) 40
Question 23
- 13/50M1
- 13/50 x 5000M1
- 1300 caoA1
- Answer: 1300
Question 24
- sample size / population = 100/250M1
- group size x sampling fraction for at least one groupM1
- 60, 16, 24 caoA1
- Answer: 60, 16, 24
Question 25
- 6/40M1
- 3/20 oeA1
- 400 x 3/20M1
- 60 caoA1
- Answer: (a) 3/20; (b) 60
Question 26
- sample size / population = 30/150M1
- group size x sampling fraction for at least one groupM1
- 10, 6, 4, 10 caoA1
- Answer: 10, 6, 4, 10
Question 27
- 10/120M1
- 1/12 oeA1
- 300 x 1/12M1
- 25 caoA1
- Answer: (a) 1/12; (b) 25
Question 28
- 14/40M1
- 7/20 x 2000M1
- 700 caoA1
- Answer: 700
Question 29
- sample size / population = 30/500M1
- group size x sampling fraction for at least one groupM1
- 9, 12, 9 caoA1
- Answer: 9, 12, 9
Question 30
- sample size / population = 40/200M1
- group size x sampling fraction for at least one groupM1
- 30, 4, 6 caoA1
- Answer: 30, 4, 6
Question 31
- 1000 x 88M1
- 88000 kg caoA1
- Answer: 88000 kg
Question 32
- sample size / population = 25/250M1
- group size x sampling fraction for at least one groupM1
- 3, 10, 12 caoA1
- Answer: 3, 10, 12
Question 33
- sample size / population = 30/200M1
- group size x sampling fraction for at least one groupM1
- 3, 12, 15 caoA1
- Answer: 3, 12, 15
Question 34
- 30/200M1
- 3/20 x 2000M1
- 300 caoA1
- Answer: 300
Question 35
- sample size / population = 100/200M1
- group size x sampling fraction for at least one groupM1
- 30, 25, 25, 20 caoA1
- Answer: 30, 25, 25, 20
Question 36
- sample size / population = 100/250M1
- group size x sampling fraction for at least one groupM1
- 8, 20, 24, 48 caoA1
- Answer: 8, 20, 24, 48
Question 37
- sample size / population = 60/180M1
- group size x sampling fraction for at least one groupM1
- 20, 20, 20 caoA1
- Answer: 20, 20, 20
Question 38
- sample size / population = 50/250M1
- group size x sampling fraction for at least one groupM1
- 10, 12, 8, 20 caoA1
- Answer: 10, 12, 8, 20
Question 39
- 500 x 65M1
- 32500 kg caoA1
- Answer: 32500 kg
Question 40
- sample size / population = 60/200M1
- group size x sampling fraction for at least one groupM1
- 12, 12, 36 caoA1
- Answer: 12, 12, 36
Question 41
- C caoB1
- Answer: C) The pupil's favourite colour
Question 42
- (a) 20/180 oe (1/9) caoB1
- (a) Answer: 20/180 (= 1/9)
- (b) 72 x 1/9, oeM1
- (b) 8 caoA1
- (b) Answer: 8
- (c) 7 cao (63 x 1/9)B1
- (c) Answer: 7
- (d) 45 x 1/9, oe (= 5)M1
- (d) 8+7+5 = 20A1
- (d) Answer: 5; 8+7+5 = 20
Question 43
- (a) 100-50-30 (= 20%), then 0.2 x 400, oeM1
- (a) 80 caoA1
- (a) Answer: 80
- (b) 50% x 40, oe (or 200/400 x 40)M1
- (b) 20 caoA1
- (b) Answer: 20
- (c) 12 cao (30% x 40)B1
- (c) Answer: 12
- (d) 20+12+8 = 40B1
- (d) Answer: 20+12+8 = 40
Question 44
- (i) False caoB1
- (i) Answer: False
- (ii) True caoB1
- (ii) Answer: True
- (iii) True caoB1
- (iii) Answer: True
- (iv) False caoB1
- (iv) Answer: False
Question 45
- (a) 50 caoB1
- (a) Answer: 50
- (b) 200+100+50+150, oeM1
- (b) 500 caoA1
- (b) Answer: 500
- (c) 100/500 x 50, oeM1
- (c) 10 caoA1
- (c) Answer: 10
- (d) 15 cao (150/500 x 50)B1
- (d) Answer: 15
- (e) 20+10+5+15 = 50B1
- (e) Answer: 20+10+5+15 = 50
Question 46
- (a) a simple random sample could by chance select very few, or no, residents from a small ward, oeB1
- (a) stratified sampling by ward guarantees every ward is represented in proportion to its population, giving a fairer picture of opinion across the whole borough, oeB1
- (a) Answer: An SRS risks under-representing small wards by chance; stratified sampling guarantees proportional representation of every ward.
- (b) it allows meaningful comparison between individual wards, because each ward is given its own known, guaranteed sample size, oeB1
- (b) this comparison would be unreliable with an SRS, where some wards might by chance have very few or no respondents at all, oeB1
- (b) Answer: Stratified sampling lets the council reliably compare results ward by ward, since every ward has a guaranteed sample size.
Question 47
- (a) there are only 84 members (numbered 01 to 84), so 91 is too large to correspond to a member, oeB1
- (a) Answer: 91 is greater than 84, so it does not match any member's register number.
- (b) member 40 has already been selected, and each member can only appear once in the sample, oeB1
- (b) Answer: Member 40 was already chosen, so selecting it again would not add a new member.
- (c) working through the digits in order, keeping only valid, non-repeated numbers of 84 or lessM1
- (c) 40, 07, 66, 12, 84 cao (in this order)A1
- (c) Answer: 40, 07, 66, 12, 84
- (d) it removes personal bias/judgement, giving every member of the stratum an equal chance of being chosen, oeB1
- (d) Answer: Random digits remove personal bias, giving every Premium member an equal chance of selection.
Question 48
- (a) membership tierB1
- (a) tier sizes differ substantially, and satisfaction may well differ by tier, so each tier needs to be represented in the sample in proportion to its size, oeB1
- (a) Answer: Membership tier; because tier sizes vary a lot and satisfaction may differ between tiers, each tier needs proportional representation.
- (b) there could be a very large number of different home towns, many containing very few members, so many strata would be tiny or empty, oeB1
- (b) whereas membership tier gives a small number of large, well-defined groups that can each be sampled from proportionally, oeB1
- (b) Answer: Home town would create huge numbers of tiny, impractical strata, whereas membership tier gives a few large, well-defined groups.
- (c) a valid variable, e.g. age (grouped into bands such as under 25, 25-44, 45+), oeB1
- (c) a valid requirement, e.g. every member's age band must be known/recorded, and the bands must cover everyone with no overlap, oeB1
- (c) Answer: Age, grouped into bands (e.g. under 25, 25-44, 45+); every member's age would need to be known and each member would need to fall into exactly one band.
Question 49
- (a) systematic - passengers/stops are chosen at a fixed, regular interval from a random starting point, oeB1
- (a) Answer: Systematic sampling
- (b) quota - interviewers fill fixed quotas for each group but choose which individuals to approach themselves (non-random selection), oeB1
- (b) Answer: Quota sampling
- (c) cluster - the population is divided into groups (depots), a small number of whole groups are randomly selected, and everyone within them is surveyed, oeB1
- (c) Answer: Cluster sampling
- (d) in stratified sampling, some members are sampled from every stratum (group), whereas in cluster sampling entire clusters are either included in full or excluded entirely, oeB1
- (d) cluster sampling therefore only ever samples from a selection of clusters, not all of them, unlike stratified sampling which always represents every group, oeB1
- (d) Answer: Stratified sampling always takes some members from every group; cluster sampling surveys everyone in only a few, randomly chosen whole groups and none from the rest.
Question 50
- (a) 60 cao (900 x 1/15)B1
- (a) Answer: 60
- (b) 12 / (1/15), oe (or 12 x 15)M1
- (b) 180 caoA1
- (b) Answer: 180
- (c) 720 x 1/15, oe (or 60-12)M1
- (c) 48 caoA1
- (c) Answer: 48
- (d) 16 cao (240 x 1/15)B1
- (d) Answer: 16
Question 51
- (a) 40/175, oeM1
- (a) 8/35 caoA1
- (a) Answer: 8/35
- (b) 40 x 8/35 (= 9.14...), oeM1
- (b) 9 (awrt)A1
- (b) Answer: 9
- (c) 10 (awrt), from 45 x 8/35 = 10.29 (2 dp)B1
- (c) Answer: 10
- (d) 6 (awrt), from 25 x 8/35 = 5.71 (2 dp)B1
- (d) Answer: 6
- (e) 9+8+7+10+6 = 40B1
- (e) Answer: 9+8+7+10+6 = 40
Question 52
- (a) 570 cao (38% of 1500)B1
- (a) Answer: 570
- (b) 38% x 150, oe (or 570/1500 x 150)M1
- (b) 57 caoA1
- (b) Answer: 57
- (c) 42 cao (28% x 150)B1
- (c) Answer: 42
- (d) 22% x 150 and 12% x 150, oeM1
- (d) 33 and 18 caoA1
- (d) 57+42+33+18 = 150 (dep on A1)B1
- (d) Answer: Full-time Humanities: 33; Part-time Science: 18; 57+42+33+18 = 150
Question 53
- (a) the trainee subtracted the sampling fraction from 60 instead of multiplying 60 by the sampling fraction (1/20), oeB1
- (a) Answer: 60 was reduced by subtracting 1/20 instead of being multiplied by 1/20.
- (b) 60 x 1/20, oeM1
- (b) 3 caoA1
- (b) Answer: 3
- (c) 3+12+30 = 45B1
- (c) Answer: 3+12+30 = 45
- (d) the overall sample would have been far larger than planned (60+12+30 = 102, not 45) and would hugely over-represent Speakers relative to their true share of delegates, oeB1
- (d) this would make the survey more expensive/time-consuming than intended and bias the overall results towards speakers' views, oeB1
- (d) Answer: The total sample would balloon to 102 (not 45) and be dominated by Speakers, making it costlier to run and biased towards speakers' opinions.
Question 54
- (a) membership numbers in each tier may have changed considerably in 6 months (e.g. members upgrading, downgrading, joining or leaving), so the original proportions may no longer reflect the current population, oeB1
- (a) using the old sample sizes would then over- or under-represent tiers relative to their current true size, oeB1
- (a) Answer: Tier sizes may have shifted since the original sample, so reusing old sample sizes could now over- or under-represent tiers compared with their true current sizes.
- (b) obtain an up-to-date list of current membership numbers in each tier (an up-to-date sampling frame), oeB1
- (b) recalculate the sampling fraction and each tier's sample size using these current numbers before selecting members, oeB1
- (b) Answer: Get up-to-date membership numbers for each tier, then recalculate the sampling fraction and each tier's sample size from these current figures.
Question 55
- (a) tutor groups are small and typically drawn from only one or two year groups, so a cluster sample of just 2 tutor groups would very likely miss one or more year groups entirely, oeB1
- (a) Answer: A cluster sample of just 2 tutor groups would very likely fail to represent all three year groups' homework habits.
- (b) 100/1000 = 1/10, oeM1
- (b) Year 9: 40A1
- (b) Year 10: 35 and Year 11: 25B1
- (b) Answer: Year 9: 40; Year 10: 35; Year 11: 25
- (c) use a random method (e.g. a random number generator) to choose 40 students from the complete Year 9 list, so every Year 9 student has an equal chance of selection, oeB1
- (c) Answer: Use a random method, such as random numbers, to pick 40 names from the full Year 9 list.
- (d) (40 x 5.4) + (35 x 6.8) + (25 x 8.1), oeM1
- (d) 656.5A1
- (d) 6.565 hours cao (awrt 6.57)A1
- (d) Answer: 6.565 hours
- (e) an unweighted average would give Year 11 the same influence as Year 9, even though Year 9 has many more students (400 compared with 250), oeB1
- (e) weighting by sample size (which reflects each year group's true share of the school) gives a fairer, more accurate estimate of the whole-school mean, oeB1
- (e) Answer: An unweighted average would let the smaller Year 11 group affect the result as much as the larger Year 9 group; weighting by sample size reflects each year's true share of the school.
Question 56
- (a) 180 cao (300-120)B1
- (a) Answer: 180
- (b) 2400/2880 x 180, oeM1
- (b) 150 caoA1
- (b) Answer: 150
- (c) 30 cao (480/2880 x 180)B1
- (c) Answer: 30
- (d) 150+30+120 = 300B1
- (d) Answer: 150+30+120 = 300
- (e) sampling Gold donors at the same overall rate (300 out of 3000, i.e. 1 in 10) would only include about 12 of the 120 Gold donors, too few to capture how varied this small, high-value group is, oeB1
- (e) including every Gold donor removes this sampling uncertainty for the group that matters most to the charity's income, oeB1
- (e) Answer: A standard proportional sample would only include about 12 Gold donors (1 in 10 of 120), too few to reliably represent this small but financially important group; including all of them avoids that risk.
Mark your answers
Check answers without an account. Some answers can be checked automatically; use the mark scheme to mark the rest.