IB Mathematics: Applications and Interpretation · Topic 4 · Statistics and probability
4.12 Designing data collection
Plan a data collection, choose a suitable sampling method, and identify and reduce sources of bias.
1/6A clear question shapes every later decision
- A clear question shapes every later decision
- A sampling frame makes random selection possible
- Stratified sampling keeps each subgroup in proportion
- Non-random methods are quicker but less trustworthy
- Bias is a systematic error that pushes results one way
- Careful design reduces bias before data collection begins
- Check your recall
- Practice: 10 questions
Part 1 of 6: A clear question shapes every later decision
A school council wants to know how students travel to school. Before asking anyone, it must decide exactly what it wants to learn. A vague aim produces data that answers nothing.
The whole group of interest is the population. The part of it that you actually measure is the sample. Planning means linking these two carefully, one step at a time.
- State the question that the data must answer.
- Identify the population and the variables to measure.
- Choose a sampling method and a sample size.
- Decide how responses will be recorded and checked.
Part 2 of 6: A sampling frame makes random selection possible
Now suppose the council wants a fair sample of its 800 students. It needs a list of everyone who could be chosen. This list is the sampling frame.
With a frame, the council can use simple random sampling. Every possible sample of the required size is equally likely. Random numbers from a calculator can make the choices.
Another option is systematic sampling. Choose a random starting point on the list, then take every th member. For 800 students and a sample of 40, .
Key terms · tap one for its definition
sampling frame
A sampling frame is a list of every member of the population from which a sample can be selected.
simple random sampling
Simple random sampling is selection in which every possible sample of the required size is equally likely to be chosen.
systematic sampling
Systematic sampling is the selection of every th member of an ordered list, starting from a randomly chosen member.

Quick check
One question on the part you just read
Easy
Show answer and explanation
Answer: B · Worked explanation
Part 3 of 6: Stratified sampling keeps each subgroup in proportion
Suppose the 800 students are 480 in Year 12 and 320 in Year 13. A simple random sample might, by chance, include too few from one year. Using stratified sampling avoids this.
Divide the population into subgroups, called strata, that share a feature. Then sample from each stratum in proportion to its size. Within each stratum, choose members at random.
Worked example
Allocating a stratified sample
- Find the population total: .
- Year 12 share: .
- Year 13 share: .
- Check the total: .
Answer: Sample 30 students from Year 12 and 20 students from Year 13.
This method works best when the strata differ on the variable being studied. Year groups may travel differently, so each must be represented. The sample then mirrors the structure of the population.
Key term
stratified sampling
Stratified sampling is the random selection of members from each subgroup in proportion to that subgroup’s size in the population.
Part 4 of 6: Non-random methods are quicker but less trustworthy
Sometimes no frame exists, or time is short. Then researchers may use methods where chance does not decide who is chosen.
In quota sampling, the interviewer fills a fixed number from each group, such as 20 men and 20 women. The interviewer picks whoever is available. In convenience sampling, the sample is simply the people easiest to reach.
Comparing sampling methods
- Method
- Simple random
- How members are chosen
- Random selection from a full list
- Main weakness
- Needs a complete sampling frame
- Method
- Systematic
- How members are chosen
- Every th member after a random start
- Main weakness
- A pattern in the list can distort it
- Method
- Stratified
- How members are chosen
- Random selection within each subgroup
- Main weakness
- Needs the subgroup sizes
- Method
- Quota
- How members are chosen
- Interviewer fills set numbers per group
- Main weakness
- Selection within groups is not random
- Method
- Convenience
- How members are chosen
- Whoever is easiest to reach
- Main weakness
- Likely to be unrepresentative
Random selection removes the chooser’s personal preferences. Quota and convenience samples are cheap and quick. Their results, however, are hard to trust because the chooser’s preferences can shape who is included.
Key terms · tap one for its definition
quota sampling
Quota sampling is the non-random selection of members until a fixed number has been reached in each specified group.
convenience sampling
Convenience sampling is the selection of whichever members of the population are easiest to reach.

Quick check
One question on the part you just read
Easy
Show answer and explanation
Answer: A · Worked explanation
Part 5 of 6: Bias is a systematic error that pushes results one way
Now consider what can go wrong. Results can show bias, which pushes them in a consistent direction. Repeating the study does not remove it.
One source is sampling bias. It occurs when the selection method makes some members more likely to be chosen than others. Asking only students at the school gate at 8 am misses those who arrive late.
Non-response is another source. It happens when chosen people do not reply. If only strongly opinionated people answer, the results overstate extreme views.
Key terms · tap one for its definition
bias
Bias is a systematic error in data collection that makes results differ from the true population value in a consistent direction.
sampling bias
Sampling bias is bias that arises when the selection method makes some members of the population more likely to be chosen than others.
Part 6 of 6: Careful design reduces bias before data collection begins
Knowing these sources, you can plan against them. Choose a method that gives every member a known chance of selection. Use a frame that covers the whole population.
A leading question hints at the expected answer, such as “Don’t you agree the canteen is poor?” Neutral wording avoids this. Anonymous responses also encourage honest answers.
Test the plan first with a small pilot study. It reveals unclear questions or impractical steps before the full investigation. Follow up non-responders where possible, and record how the data were collected.
Key terms · tap one for its definition
leading question
A leading question is a question worded so that it suggests or encourages a particular answer.
pilot study
A pilot study is a small trial run of a data collection plan, used to find problems before the full study.
Check your recall[16 marks]
6 written questions. Answer each one in your head or on paper, then open the mark scheme.
- 1.State[2 marks]
State two decisions that must be made when planning a data collection.
Show mark schemeHide mark scheme
Model answer
Decide which question the data must answer, and choose a sampling method and sample size.
- 1 mark for stating a valid decision, such as identifying the population or the variables to measure
- 1 mark for stating a second valid decision, such as choosing the sampling method, sample size or recording method
- 2.Calculate[3 marks]
A college has 540 day students and 360 boarders. Calculate how many students from each group a stratified sample of 60 should contain.
Show mark schemeHide mark scheme
Model answer
Total . Day students: . Boarders: .
- 1 mark for finding the population total of 900
- 1 mark for obtaining 36 day students
- 1 mark for obtaining 24 boarders
- 3.Distinguish[3 marks]
Distinguish between quota sampling and stratified sampling.
Show mark schemeHide mark scheme
Model answer
Both divide the population into groups and sample from each. In stratified sampling, members within each group are chosen at random in proportion to group size. In quota sampling, the interviewer fills fixed numbers with whoever is available, so selection is not random.
- 1 mark for stating that both use groups within the population
- 1 mark for stating that stratified sampling selects members at random within each group, in proportion to its size
- 1 mark for stating that quota sampling fills fixed numbers non-randomly with whoever is available
- 4.Explain[2 marks]
Explain why increasing the sample size does not remove sampling bias.
Show mark schemeHide mark scheme
Model answer
Sampling bias comes from the selection method making some members more likely to be chosen. A larger sample taken by the same method repeats the same favouring, so results still differ from the true value in the same direction.
- 1 mark for stating that bias is caused by the selection method, not by the sample size
- 1 mark for explaining that a larger sample from the same method keeps the same consistent error
- 5.Describe[3 marks]
Describe how a systematic sample of 30 people could be taken from a list of 600 people.
Show mark schemeHide mark scheme
Model answer
Calculate . Choose a random starting person among the first 20 on the list. Then select every 20th person after that.
- 1 mark for calculating
- 1 mark for choosing a random starting point
- 1 mark for selecting every 20th person from that start
- 6.Discuss[3 marks]
Discuss the use of convenience sampling to find the mean spending of all visitors to a shopping centre.
Show mark schemeHide mark scheme
Model answer
It is quick and cheap, and needs no sampling frame. However, it is not random. Visitors at one entrance or time may spend differently from others, causing sampling bias, so conclusions about all visitors may be unreliable.
- 1 mark for giving an advantage, such as speed, low cost or no frame needed
- 1 mark for giving a disadvantage, such as non-random selection causing sampling bias
- 1 mark for reaching a judgement that results may not represent all visitors

Practice this topic
10 quick-fire multiple-choice questions, easy to hard. Tap an answer to see if you're right and the worked explanation. (The written questions with mark schemes are under “Check your recall” above.)
0 right0 of 10 answered
Includes the 2 quick checks in the note above.
EasyIdentify[1 mark]
Show answer and explanation
Answer: C · Worked explanation
MediumDetermine[1 mark]
Show answer and explanation
Answer: B · Worked explanation
MediumCalculate[2 marks]
Show answer and explanation
Answer: D · Worked explanation
MediumIdentify[1 mark]
Show answer and explanation
Answer: A · Worked explanation
MediumIdentify[1 mark]
Show answer and explanation
Answer: C · Worked explanation
HardExplain[2 marks]
Show answer and explanation
Answer: B · Worked explanation
HardDistinguish[2 marks]
Show answer and explanation
Answer: A · Worked explanation
HardCalculate[2 marks]
Show answer and explanation
Answer: C · Worked explanation