All subjects · Geography · WGE02 Geographical Investigations

Sampling strategies and sample size

Edexcel IAL Geography · WGE02 Geographical Investigations

You can rarely measure everything, so you sample. How you choose the sample decides whether results are representative and reliable (Section B and C questions often ask about it).

Practise this in the app: instant marking and a fix for every mistake. Free, no email needed to start.

Practise this lesson

Four sampling methods

Random: every point or person has an equal chance of selection (random number tables or a random number generator for grid coordinates). Avoids bias but may miss parts of the area by chance and can be impractical on the ground.
Systematic: samples at regular intervals (every 20 m along a transect, every 10th person). Easy and gives even coverage; may miss or over-represent patterns that repeat at that interval. The June 2025 mark scheme identified sites at regular spacing as systematic.
Stratified: the population is divided into sub-groups (strata), such as dune zones or locals and visitors, and each is sampled in proportion to its size, or equally. Makes sure every group is represented.
Opportunistic (opportunity): sampling whoever or whatever is available. Quick, but biased.

They can be combined: stratified systematic (regular intervals within each zone) or stratified random.

Point, line and area sampling

Point: individual spots (a grid of points for land use).
Line (transect): samples along a line across a change, such as from the sea inland across dunes, or from a city centre outwards. A belt transect samples continuously; an interrupted belt transect samples at intervals.
Area (quadrat): a square frame (e.g. 1 m x 1 m) to estimate percentage cover or count species.

The June 2024 mark scheme credited an advantage of a sampling procedure such as equal numbers of locals and visitors (stratified) with a sample size of 20, and noted that 'accurate', 'valid' or 'reliable' earn credit only when developed.

Sample size

Larger samples (January 2024 mark scheme): include more viewpoints so results are more valid; allow fuller statistical analysis and more accurate averages; make anomalies easier to identify. 'More data' alone earned no credit. Statistical tests also need minimum sizes: Spearman's rank is usually used with at least 10 pairs of data.

Method	Best for	Main weakness
Random	Avoiding bias	Uneven coverage; impractical
Systematic	Even coverage along transects	May miss repeating patterns
Stratified	Representing known sub-groups	Needs knowledge of group sizes
Opportunistic	Speed when time is short	Biased, unrepresentative

Worked example

A town has about 600 residents and 400 visitors on the seafront on a survey day. You plan 50 questionnaires using proportional stratified sampling. How many residents should you ask?

Residents' share = 600 / (600 + 400) = 0.6.
0.6 x 50 = 30.

Answer: 30 residents and 20 visitors.

Key idea: Random (unbiased), systematic (regular intervals), stratified (sub-groups represented), opportunistic (quick but biased). Point, line (transect) and area (quadrat) sampling. Larger samples improve validity, averages, anomaly detection and statistical testing.

Check yourself

A dune system has three zones covering 50%, 30% and 20% of the transect. In a proportional stratified sample of 40 quadrats, how many should be in the 30% zone?
Show answer
12. 0.30 x 40 = 12.
A 240 m transect is sampled systematically every 20 m, starting at 0 m and including 240 m. How many sites is that?
Show answer
13. 240 / 20 = 12 intervals, so 13 sites (including both ends).
Sites at 0, 50, 100, 150, 200 and 250 m from a city centre were sampled. What sampling method was used?
- Systematic
- Random
- Stratified
- Opportunistic
Show answer
Systematic. Regularly spaced sites indicate systematic sampling (June 2025 mark scheme).
Why might you use stratified sampling to interview locals and visitors about a coastal scheme?
- To interview whoever walks past
- To pick people at regular intervals only
- To avoid having to count anyone
- To make sure the views of each group are represented
Show answer
To make sure the views of each group are represented. Strata guarantee each group is included.
The main disadvantage of opportunistic sampling is that
- it takes too long
- it is likely to be biased and unrepresentative
- it needs random number tables
- it requires a transect
Show answer
it is likely to be biased and unrepresentative. Only the available or willing are sampled.
A quadrat is used for
- measuring wind speed
- recording noise
- counting traffic
- area sampling, such as estimating percentage plant cover
Show answer
area sampling, such as estimating percentage plant cover. It defines a fixed area.
Explain one advantage of using a large sample size in primary data collection.
Show answer
A larger sample includes more people or sites, so results are more representative and valid, averages are more accurate, anomalies are easier to spot and statistical tests can be carried out reliably.. Name the benefit and link it to size.

Practise with instant feedback

Sources

Last updated 8 November 2026. Spotted a mistake? Tell us.