01 / Collecting data and samplingEvery board · Both tiers
In statistics, the population is:
- Athe people who answered your survey
- Ba list of names to choose from
- Ceveryone or everything you want to find out about
- Dthe number of people in a country
Show the answer
Answer: C everyone or everything you want to find out about
The population is the whole group the enquiry is about, stated in full, such as “all Year 7 pupils at the school”. It depends on the question, so it is not always people in a country.
Checked against assets.publishing.service.gov.uk
02 / Collecting data and samplingEvery board · Both tiers
A head teacher wants to know what all 900 pupils at her school think of the canteen. She asks 60 of them.
What is the population?
- AAll 900 pupils at the school
- BAll secondary pupils in England
- CThe 60 pupils she asks
- DThe canteen staff
Show the answer
Answer: A All 900 pupils at the school
The population is the whole group of interest (all 900 pupils); the sample is the part actually asked (60 pupils). Giving the sample as the population is a common slip.
Checked against qualifications.pearson.com
03 / Collecting data and samplingEvery board · Both tiers
A census collects data from:
- Aevery member of the population
- Ba random sample of the population
- Cthe people easiest to reach
- Done person from each group
Show the answer
Answer: A every member of the population
A census includes every member of the population, so nobody is left out. That makes it complete, but slow and costly, especially for a large population.
Checked against assets.publishing.service.gov.uk
04 / Collecting data and samplingEvery board · Both tiers
A factory wants to know how long its light bulbs last. Each bulb tested is left on until it fails.
Why would the factory use a sample, not a census?
- AA sample is always more accurate than a census
- BTesting destroys the items tested
- CBulbs are qualitative data
- DA census needs a hypothesis
Show the answer
Answer: B Testing destroys the items tested
If testing destroys the item, a census would leave nothing to sell. Destructive testing is a key reason for sampling, and one pupils often miss.
Checked against qualifications.pearson.com
05 / Collecting data and samplingEvery board · Both tiers
Why do researchers often use a sample instead of a census?
- AIt removes all bias
- BIt always gives exact answers
- CIt is quicker and cheaper
- DIt includes everybody
Show the answer
Answer: C It is quicker and cheaper
A sample saves time and money and gives less data to process. The risk is that it does not represent the population, so it must be chosen carefully.
Checked against assets.publishing.service.gov.uk
06 / Collecting data and samplingEvery board · Both tiers
A sampling frame is:
- Aa chart showing who was in the sample
- Bthe time the survey takes
- Cthe size of the sample
- Da list of the population to choose the sample from
Show the answer
Answer: D a list of the population to choose the sample from
A sampling frame is the list of members the sample is chosen from, such as a school register or the electoral register. Examiners say very few pupils know this term.
Checked against cdn.sanity.io
07 / Collecting data and samplingEvery board · Both tiers
In a simple random sample:
- Aevery tenth member is chosen
- Bevery member has the same chance of being chosen
- Cpeople are picked to fill set numbers
- Dthe researcher picks typical members
Show the answer
Answer: B every member has the same chance of being chosen
In a simple random sample every member of the population is equally likely to be picked. Random number tables, a calculator, dice or names drawn from a hat can be used.
Checked against assets.publishing.service.gov.uk
08 / Collecting data and samplingEvery board · Both tiers
When choosing a sample with random numbers, what do you do with a number that comes up twice?
- AStart the sample again
- BChoose that member twice
- CRound it to the nearest ten
- DIgnore it and take another
Show the answer
Answer: D Ignore it and take another
You need a full list (the sampling frame), with each member numbered. Then random numbers pick the members; repeats are ignored so nobody is chosen twice.
Checked against qualifications.pearson.com
09 / Collecting data and samplingEvery board · Both tiers
Members of a club are numbered 001 to 250. The random number 317 comes up.
What should you do with 317?
- AStop and use a smaller sample
- BChoose member 31 instead
- CIgnore it and use the next number
- DChoose member 250 instead
Show the answer
Answer: C Ignore it and use the next number
There is no member 317, so the number is out of range and is ignored. Changing it to another member would make some members more likely to be picked than others.
Checked against qualifications.pearson.com
10 / Collecting data and samplingEvery board · Both tiers
In systematic sampling you:
- Apick a random start, then every kth member
- Bpick every member in turn
- Cpick the first k members
- Dpick whoever happens to be nearest
Show the answer
Answer: A pick a random start, then every kth member
Systematic sampling takes every kth member from a list, starting at a random place between 1 and k. Without the random start, members later in the first block could never be picked.
Checked against qualifications.pearson.com
11 / Collecting data and samplingEvery board · Both tiers
Houses on an estate are listed in order, and every 10th house on the list is a large corner house.
Why could a systematic sample taking every 10th house be biased here?
- AThe houses are not numbered
- BThe sample would be too large
- CIt could pick only corner houses, or none
- DSystematic samples are never random
Show the answer
Answer: C It could pick only corner houses, or none
If the list has a pattern that repeats at the same interval as the sample, a systematic sample can pick only one kind of member. This is the main weakness of systematic sampling.
Checked against qualifications.pearson.com
12 / Collecting data and samplingEvery board · Both tiers
In a stratified sample, the population is split into groups and then:
- Aone group is chosen at random
- Beach group is sampled in proportion to its size
- Conly the largest group is sampled
- Dthe same number is taken from each group
Show the answer
Answer: B each group is sampled in proportion to its size
Stratified sampling makes sure each group is represented in proportion to its size. Taking equal numbers from each group, whatever its size, is a slip both boards report.
Checked against cdn.sanity.io
13 / Collecting data and samplingEvery board · Both tiers
Splitting a population into groups, such as age groups, before sampling is called:
- Arandomisation
- Bstratification
- Ccleaning
- Dinterpolation
Show the answer
Answer: B stratification
Stratification is the first step: split the population into groups (strata). Then a random sample is taken from each group, which pupils often forget to say.
Checked against aqa.org.uk
14 / Collecting data and samplingEvery board · Both tiers
In quota sampling, the interviewer:
- Apicks every kth person on a list
- Basks every member of the population
- Cchooses people until each group’s number is filled
- Dpicks people from a list using random numbers
Show the answer
Answer: C chooses people until each group’s number is filled
In quota sampling the interviewer fills set numbers from each group, for example 20 men and 20 women, choosing whoever they like. It is quick and needs no list, but it is not random.
Checked against assets.publishing.service.gov.uk
15 / Collecting data and samplingEvery board · Both tiers
To find out what shoppers think of a new shopping centre, a student asks the first 30 people she meets at the entrance.
What type of sample is this?
- AA stratified random sample
- BA systematic sample
- CA random sample
- DA convenience sample
Show the answer
Answer: D A convenience sample
A convenience (or opportunity) sample uses whoever is easy to reach. It is quick and cheap but likely to be biased: the people at one entrance at one time may not be typical. Write “convenience”, not “convenient”.
Checked against cdn.sanity.io
16 / Collecting data and samplingEvery board · Both tiers
In judgement sampling, members are chosen:
- Aat regular intervals from a list
- Bin proportion to group sizes
- Cat random from a list of the whole population
- Dby a researcher who picks those thought typical
Show the answer
Answer: D by a researcher who picks those thought typical
In a judgement sample the researcher uses their own knowledge to choose members they think represent the population. It can be quick and use expert knowledge, but their opinion may introduce bias.
Checked against assets.publishing.service.gov.uk
17 / Collecting data and samplingEvery board · Both tiers
A sample is biased when:
- Ait is smaller than the population
- Bit is chosen at random
- Cit does not fairly represent the population
- Dit contains numbers
Show the answer
Answer: C it does not fairly represent the population
Bias means the sample leans one way, so results are pushed away from the truth about the population. It can come from who is asked, how, when, and who chooses to answer.
Checked against assets.publishing.service.gov.uk
18 / Collecting data and samplingEvery board · Both tiers
To find how often people in a town use buses, a survey is done at the bus station.
Why is this sample biased?
- ABus users are more likely to be asked
- BBuses run at set times
- CThe sample is too random
- DThe data is qualitative
Show the answer
Answer: A Bus users are more likely to be asked
People at a bus station are much more likely to use buses than people in the town as a whole, so the results would overstate bus use. Where and when you ask matters.
Checked against assets.publishing.service.gov.uk
19 / Collecting data and samplingEvery board · Both tiers
Which change would make a survey sample less likely to be biased?
- AAsk only people you know
- BAsk people at one place and time
- CLet people choose to take part
- DChoose people at random from a list
Show the answer
Answer: D Choose people at random from a list
Random selection gives every member the same chance, so the sample is less likely to be biased. In the exam say “less likely to be biased”; “it avoids bias” is too strong to earn the mark.
Checked against cdn.sanity.io
20 / Collecting data and samplingEvery board · Both tiers
Why is a larger sample usually better?
- AIt removes all bias
- BIt is cheaper to collect
- CThe results are more reliable
- DIt is quicker to collect
Show the answer
Answer: C The results are more reliable
Larger samples give estimates that vary less and are more likely to be repeated, so they are more reliable. But they cost more, and a large biased sample is still biased.
Checked against cdn.sanity.io
21 / Collecting data and samplingEvery board · Both tiers
Results are reliable if:
- Arepeating the method gives similar results
- Bthey support the hypothesis you started with
- Cthey come from a census
- Dan expert collected them
Show the answer
Answer: A repeating the method gives similar results
Reliable means you would get similar results if you did it again. Valid means it measures what it claims to. Data is not reliable just because an expert collected it.
Checked against cdn.sanity.io
22 / Collecting data and samplingEvery board · Both tiers
A questionnaire asks: “How much time do you spend on homework?” The boxes are: 0 to 1 hour, 1 to 2 hours, 2 to 3 hours.
Which of these is a fault with the question?
- AIt is a closed question
- BIt asks about time
- CIt does not say per day or per week
- DIt is about homework
Show the answer
Answer: C It does not say per day or per week
A good question gives a time frame, such as “per week”. Response boxes must not overlap (1 hour is in two boxes here) and must cover every answer, including an open top box such as “more than 3 hours”.
Checked against filestore.aqa.org.uk
23 / Collecting data and samplingEvery board · Both tiers
Response boxes for a closed question should:
- Acover every possible answer without overlapping
- Balways have exactly three boxes
- Coverlap so nobody is left out
- Dleave out zero
Show the answer
Answer: A cover every possible answer without overlapping
Every possible answer must fit in exactly one box. That means no gaps, no overlaps, a box for 0 where it can happen, and an open box at the top.
Checked against qualifications.pearson.com
24 / Collecting data and samplingEvery board · Both tiers
What is an advantage of an open question?
- AThe answers are easy to analyse
- BThe answers fit set categories
- CPeople can give detailed answers
- DIt always gets a very quick reply
Show the answer
Answer: C People can give detailed answers
Open questions let people answer in their own words, which gives detail but is hard to analyse. Closed questions give set choices, which are quick to answer and easy to group.
Checked against assets.publishing.service.gov.uk
25 / Collecting data and samplingEvery board · Both tiers
A survey question reads: “Don’t you agree that school starts too early?”
What is wrong with this question?
- AIt asks for a number
- BIt is about the school day
- CIt is an open question
- DIt is a leading question
Show the answer
Answer: D It is a leading question
A leading question suggests the answer the writer wants, so people are more likely to agree. A neutral version would be “What do you think of the time school starts?” with clear options.
Checked against assets.publishing.service.gov.uk
26 / Collecting data and samplingEvery board · Both tiers
Why might a question about how much money someone earns give unreliable answers?
- AMoney is qualitative data
- BPeople may not answer honestly
- CEarnings cannot be put in groups
- DThe answers would all be the same
Show the answer
Answer: B People may not answer honestly
People may refuse, or give the answer they think sounds better, when a question is sensitive. Anonymous answers and grouped response boxes make honest answers more likely.
Checked against assets.publishing.service.gov.uk
27 / Collecting data and samplingEvery board · Higher only
Each person secretly flips a coin. Heads: they answer “Yes”. Tails: they answer the sensitive question truthfully.
Why is this method used?
- AIt makes the sample bigger
- BIt picks people at random
- CNobody knows if a person’s “Yes” is true
- DIt removes the need for a sample
Show the answer
Answer: C Nobody knows if a person’s “Yes” is true
The random response technique protects people: a “Yes” might have been forced by the coin, so people can answer a sensitive question honestly. It is about honesty, not about choosing people at random.
Checked against qualifications.pearson.com
28 / Collecting data and samplingEvery board · Both tiers
What is an advantage of a face-to-face interview over a postal questionnaire?
- AIt reaches more people for less money
- BThe interviewer can explain unclear questions
- CIt is always cheaper
- DThe interviewer cannot affect the answers
Show the answer
Answer: B The interviewer can explain unclear questions
Interviews usually get more replies and let questions be explained, but they cost more and the interviewer’s manner can sway answers (interviewer bias). Postal or online questionnaires are cheaper but get fewer replies.
Checked against assets.publishing.service.gov.uk
29 / Collecting data and samplingEvery board · Both tiers
Many people do not return a postal questionnaire. Which would help most?
- ARun a pilot study
- BMake the questionnaire longer
- CSend a reminder or follow up
- DUse more leading questions
Show the answer
Answer: C Send a reminder or follow up
People who reply may care more about the topic than people who do not, so the results lean their way. Following up, keeping it short or offering a small reward helps. A pilot study does not fix non-response afterwards.
Checked against qualifications.pearson.com
30 / Collecting data and samplingEvery board · Both tiers
A pilot study is:
- Aa group that gets no treatment
- Ba list of the population
- Cthe main survey of the whole population at once
- Da small trial run before the main survey
Show the answer
Answer: D a small trial run before the main survey
A pilot study tries the questionnaire or method on a few people first. It shows up unclear questions, missing response boxes and timing problems, so they can be fixed before the main survey.
Checked against qualifications.pearson.com
31 / Collecting data and samplingEvery board · Both tiers
What is an advantage of a laboratory experiment?
- ANobody knows they are being tested
- BExtraneous variables can be controlled
- CIt happens in a natural setting
- DIt needs no planning
Show the answer
Answer: B Extraneous variables can be controlled
A laboratory experiment gives control and can be repeated, but the setting is artificial. A field experiment happens in a real setting, which is more natural but harder to control.
Checked against assets.publishing.service.gov.uk
32 / Collecting data and samplingEvery board · Both tiers
Researchers compare road accidents in a town before and after the council lowered the speed limit.
What type of experiment is this?
- AA simulation
- BA pilot study
- CA natural experiment
- DA laboratory experiment
Show the answer
Answer: C A natural experiment
In a natural experiment the change happens anyway (here, the council lowered the limit) and researchers study its effect. They cannot control the explanatory variable or the extraneous ones.
Checked against assets.publishing.service.gov.uk
33 / Collecting data and samplingEvery board · Higher only
In an experiment, a control group is a group that:
- Acollects the data
- Bgets a double dose of the treatment
- Cchooses its own treatment
- Ddoes not get the treatment, for comparison
Show the answer
Answer: D does not get the treatment, for comparison
A control group is as similar as possible to the treated group but does not get the treatment. Comparing the two shows whether the treatment made a difference.
Checked against assets.publishing.service.gov.uk
34 / Collecting data and samplingEvery board · Higher only
To test a new revision app, one class uses the app and another similar class does not.
What is the class without the app called?
- AThe pilot group
- BThe quota
- CThe control group
- DThe sampling frame
Show the answer
Answer: C The control group
The class without the app is the control group. If the classes are alike in ability and teaching, a difference in results is more likely to come from the app than from something else.
Checked against assets.publishing.service.gov.uk
35 / Collecting data and samplingEvery board · Both tiers
Which of these is part of cleaning data?
- ARemoving every value above the mean
- BChanging results to fit the hypothesis
- CRemoving an impossible value, such as an age of 250
- DDeleting the largest value, even though it is genuine
Show the answer
Answer: C Removing an impossible value, such as an age of 250
Cleaning means finding and fixing problems before analysis: impossible values, missing values, mixed units or formats and repeated entries. Genuine values are kept, even extreme ones.
Checked against assets.publishing.service.gov.uk
36 / Collecting data and samplingEvery board · Both tiers
A table of heights has some values in metres, such as 1.62, and others in centimetres, such as 158.
What should be done before using the data?
- ADelete all the values given in metres
- BChange them all to the same units
- CFind the mean straight away
- DUse only the largest values
Show the answer
Answer: B Change them all to the same units
Mixed units make values impossible to compare. Converting every height to the same unit is part of cleaning; examiners want the actual problem named, not just “remove errors”.
Checked against qualifications.pearson.com
37 / Collecting data and samplingEvery board · Both tiers
A spreadsheet cell holds “£12.50” as text instead of the number 12.50.
Why can this cause a problem?
- AIt is an outlier and must be deleted
- BThe software may not treat it as a number
- CIt is too large for the spreadsheet to store
- DIt is a leading question
Show the answer
Answer: B The software may not treat it as a number
Symbols, spaces or words typed into number cells can stop a spreadsheet using them in calculations. Cleaning the data removes them so every value is read correctly.
Checked against assets.publishing.service.gov.uk
38 / Collecting data and samplingEvery board · Both tiers
In a survey, some people left the “age” question blank.
What is this problem called?
- AExtrapolation
- BMissing values
- CStratification
- DOutliers
Show the answer
Answer: B Missing values
Blank answers are missing values. They must be dealt with when cleaning the data, for example by leaving that person out of the age analysis, and the number missing should be reported.
Checked against cdn.sanity.io
39 / Collecting data and samplingEvery board · Both tiers
Which organisation runs the census in England and Wales?
- AThe Bank of England
- BThe Royal Statistical Society
- CThe Office for National Statistics
- DThe Department for Education (DfE)
Show the answer
Answer: C The Office for National Statistics
The Office for National Statistics (ONS) runs the census in England and Wales every 10 years; the latest was Census 2021. Its tables are free secondary data for enquiries.
Checked against ons.gov.uk
40 / Collecting data and samplingEvery board · Both tiers
The mean of a random sample can be used to:
- Aestimate the mean of the population
- Bchoose the sampling frame
- Cremove bias from the sample
- Dfind the exact population mean
Show the answer
Answer: A estimate the mean of the population
A sample statistic, such as a sample mean or proportion, estimates the same figure for the whole population. It is only an estimate, and it is better with a larger random sample.
Checked against assets.publishing.service.gov.uk