4.1 Populations, samples and data collection
Studying this course? Keep your progress in one place.
Key facts
- A population is the whole group; a sample is the part you collect data from. In a random sample every member has an equal chance.
- Discrete data are counted; continuous data are measured.
- Five techniques: simple random, systematic, stratified, quota, convenience. Know how each works and one strength and weakness.
- Stratified numbers: .
- Outlier boundaries: and . Outliers can be genuine or errors; missing values are excluded, never set to 0.
Populations, samples and data types
The population is everyone you want to know about; the sample is who you actually measure.
The population is the whole group you want to know about. A sample is the part you collect data from. In a random sample every member has an equal chance of being chosen, and choices are independent.
Discrete data are counted and take particular values (people in a household, shoe sizes). Continuous data are measured and take any value in a range (time, mass, electricity use).
Continuous data are always recorded to some accuracy, but that does not make them discrete. Age is continuous, though written in whole years.
| Discrete | Continuous | |
|---|---|---|
| How obtained | Counted | Measured |
| Possible values | Particular values only | Any value in a range |
| Examples | Goals, shoe sizes | Time, mass, age |
Discrete
- How obtained:
- Counted
- Possible values:
- Particular values only
- Examples:
- Goals, shoe sizes
Continuous
- How obtained:
- Measured
- Possible values:
- Any value in a range
- Examples:
- Time, mass, age
Is the time taken to run 100 m discrete or continuous?
Sampling techniques
Simple random, systematic and stratified choose randomly; quota and convenience do not.
- Simple random: number every member from a sampling frame and pick random distinct numbers.
- Systematic sampling: pick every th member after a random start, . Biased if the list has a repeating pattern.
- Stratified: split into strata and sample randomly from each in proportion to its size.
- Quota sampling: fill a set number from each group, choosing non-randomly. Quick, but open to bias.
- Convenience sampling: use whoever is easiest to reach. Cheap but usually biased.
Random
- Simple random
- Systematic
- Stratified
Not random
- Quota
- Convenience
- 1
Get a sampling frame
a list of every member
- 2
Number the members
1 to N
- 3
Generate random numbers
with a calculator or software
- 4
Select the matching members
ignore repeats until the sample is full
- Stratified numbers
Worked example
There are 330 DP1 and 270 DP2 students. Take a sample of 40, stratified by year.
- 1
Population size: .
- 2
DP1: .
- 3
DP2: .
A school has 600 students and wants a systematic sample of 40. What is ?
Bias and reliability
Name who is over- or under-represented, why it matters for the variable, and which way the estimate is pushed.
A sampling method is biased if it tends to over- or under-represent part of the population, so results are systematically wrong.
When explaining bias, name who is left out or over-represented and why that matters for the variable being measured. 'It is not random' alone earns little credit. Link every comment to the context.
The reliability of a source depends on who collected it, how, and why. A sample at a railway station at 8 a.m. says little about people who drive or work nights.
Stratified sampling helps when the variable differs between groups.
Weak comment
- It is not random
- The sample is too small
Strong comment
- Names who is over-represented
- Says why it matters for the variable
- Gives the direction of the bias
A survey on healthy eating is given only to people leaving a gym. Which is the best comment?
Outliers, errors and missing data
A value more than beyond a quartile is an outlier; decide in context whether to keep it.
An outlier is a value more than below or above .
Being an outlier does not mean a value is wrong. A genuine extreme (a very tall basketball player) should be kept. An error (a negative electricity reading, a typing mistake) should be corrected if possible, or removed.
Missing data should be left out of calculations, not replaced by 0, which would distort the mean. Report how many values were excluded.
- Lower boundary
- Upper boundary
Worked example
and . Find the outlier boundaries.
- 1
and .
- 2
Lower: . Upper: .
and . Is a value of 95 an outlier?
Try an exam question
A school has 600 students: 420 in the lower school and 180 in the upper school. A sample of 30 students is taken, stratified by school. (a) State one advantage of stratified sampling. (b) Find the number of students sampled from each part of the school. (c) State one disadvantage of convenience sampling.
[4 marks]
- [1]Each group is represented in proportion to its size, so the sample reflects the population.
- [1]Lower school: .
- [1]Upper school: .
- [1]Convenience sampling is usually biased, because it only includes those easiest to reach.
That's the notes covered.
Carry on to the next subtopic.
Try a question
The sample is stratified by year group, in proportion to the size of each year group. Find the number of DP1 students in the sample.
1 mark, marked against the mark scheme
The sample is stratified by year group, in proportion to the size of each year group. Find the number of DP1 students in the sample.
1 mark, marked against the mark scheme