Sampling is how you measure a part of an area fairly when you cannot measure all of it. This lesson compares three approaches on one original survey and shows how to choose.
It supports the investigation route in designing a small ethical geographical investigation. Once you have chosen, choosing a sampling method and explaining its limitation shows how to write about it.
What do the three approaches mean?
| Approach | How points are chosen | Main risk |
|---|---|---|
| Random | By chance, such as drawn numbers | Points may bunch together and miss part of the area |
| Systematic | At equal spacing from a start point | A regular pattern in the area can match the spacing |
| Stratified | In proportion to each zone or type | Zones must be known and measured first |
The next section puts one survey through all three.
Worked example: litter along a promenade
This is a worked example. A class wants to describe litter on a 600 m promenade. The first 200 m is beside a food court, and the remaining 400 m is beside a garden. The class has time for 12 points.
Systematic: start at 50 m and measure every 50 m, giving points at 50, 100, 150 and so on to 600, which is 12 points. The spacing is fair along the whole length.
Random: draw 12 distances between 0 and 600 m from numbers on slips. By chance, 7 of them might land in the first 200 m, which would overstate the food court.
Stratified: the food court is 200 ÷ 600 = one third of the length, and the garden is two thirds. Of 12 points, 12 ÷ 3 = 4 go in the food court strip and 8 in the garden strip. This keeps the sample in line with the real proportions.
The mistake that loses marks
The common slip is to name an approach with no reason: “I used random sampling.” The marker cannot tell whether it fits the question.
| Weak | Strong |
|---|---|
| I used random sampling. | I used stratified sampling because the food court and garden strips differ and I wanted both fairly represented. |
| Systematic is the best. | Systematic sampling every 50 m covers the whole promenade, but a regular pattern such as bins every 50 m could bias it. |
A strong answer links the approach to the question and states the weakness in one sentence.
Which approach suits which question?
If the question asks “what is litter like along the whole promenade?”, systematic covers the length evenly. If it compares two zones, stratified keeps both fair. If you have no prior knowledge of the area and want to avoid choosing points yourself, random removes your own bias.
Whichever you choose, the number of points and the time of day should be stated. Set out your choice and its weakness using the graph evidence and fair-comparison lab.
Check yourself
Original scenario: a fictional stream is 900 m long. The upper 300 m runs through forest and the lower 600 m through farmland. A class has time for 12 points. How many points go in each part under stratified sampling, and what is one risk of systematic sampling every 75 m?
Answer
Forest share = 300 ÷ 900 = one third, so 12 ÷ 3 = 4 points. Farmland share = 600 ÷ 900 = two thirds, so 8 points.
Systematic sampling every 75 m gives 900 ÷ 75 = 12 points. One risk: if a farm drain enters the stream every 150 m, every second point would fall at a drain, so the sample would be biased.
What to study next
Go on to choosing a sampling method and explaining its limitation, then see writing a research question and presenting data clearly.
For a teacher to check your sampling choices against a new question, see online one-to-one Geography tuition.