Geographical skills · GCSE Geography
Sampling methods in geography
GCSE Geography skills revision on random, systematic and stratified sampling: how to justify a method against a hypothesis, and why convenience sampling is a weakness you should admit rather than disguise.
Sampling is the method, not an extra word. Random is equal chance, systematic is every nth, stratified is fair shares of environments — convenience is where the coach parked.
The important bits
What you need to know
- 1
A sample is a subset of a population (pebbles, pedestrians, houses, vegetation quadrats) used because you cannot measure everything. The enquiry question decides what the population is. If you cannot define the population, you cannot defend the sample.
- 2
Random sampling: every location or person in the defined population has an equal chance of selection (random-number tables, generator, hat). It reduces researcher bias. It can miss a rare but important environment, and it is awkward on a dangerous river.
- 3
Systematic sampling: a regular interval — every 10 m along a transect, every fifth pedestrian, every tenth house. It is fast and gives even spatial coverage. It can coincide with a hidden pattern (lamp posts, pool-riffle, terrace housing) and create a false rhythm.
- 4
Stratified sampling: the sample is split in proportion to sub-groups or environments — 40% of points on the upper beach if 40% of the beach area is upper shore; or interviews matching age bands. It stops you over-studying the easy sand by the steps.
- 5
Justification is the mark: name the technique, link it to the hypothesis, and say what bias it avoids. “We used random sampling because it is random” scores nothing.
- 6
Convenience (opportunity) sampling: you measure where it is safe, dry, or close to the minibus. It is common in school fieldwork. It is weaker. Honesty plus a limitation scores higher than calling a teacher-placed pole “random”.
- 7
Sample size and reliability: five pebbles at one site is a story, not a pattern. Reliability is whether a repeat would look similar; accuracy is closeness to the true value. Sampling design hits reliability first.
- 8
Ethics and access shape samples: private land, tides, traffic, and who will stop for a clipboard. A Saturday-afternoon questionnaire is not “the town”. State the frame, then the flaw. That is evaluation, not a confession that the trip was pointless.
Quotations worth analysing
Short evidence. Real method.
“Random means equal chance, not “we wandered about”.”
Examiners have read hundreds of diaries. If you did not use random numbers, do not write random. Name what you actually did.
“Systematic sampling is even coverage with a risk of hitting a hidden beat.”
Every 10 m on a river is defensible for a downstream-change hypothesis. Admit the pool-riffle coincidence risk.
“Convenience sampling is a limitation, not a secret.”
Access and safety are real. Calling them random is the error. Propose stratified or longer transects as the improvement.
Go deeper
Justify, do not brandish the word
If you systematically sample a river every 10 metres you may miss a pool-riffle sequence that random points would catch — or you may neatly catch the downstream trend your Bradshaw hypothesis needs. Stratified sampling on a beach (equal or proportional points on upper, mid, lower shore) stops you over-studying the easy sand nearest the steps. In the write-up, name the technique, justify it against the hypothesis, and admit the bias of access. Examiners are tired of “we used random sampling” from students who stood where the ranging poles were planted. Honesty plus a sensible limitation scores higher than a textbook word used inaccurately. A 4-mark method question is won by: population defined, method named, reason linked to theory, one practical constraint.
Go deeper
Random is a design, not a vibe
True random sampling needs a sampling frame: a numbered grid over an OS map, a list of houses, a random-number source. Throwing a quadrat over your shoulder is closer to haphazard than random, and it still avoids the bog you do not want to enter. Random designs reduce the researcher choosing photogenic pebbles or friendly shoppers. They can scatter points into a weir, a dual carriageway, or a private garden — hence risk assessment belongs in the method, not in a different essay. If safety forced you off random points, say so and switch to systematic along a safe bank. That decision is geographical. Pretending otherwise is how candidates lose the evaluation level.
Go deeper
Stratify when the environment is already striped
Beaches, CBDs, and woodlands come in zones. If 70% of your pedestrian study area is a precinct and 30% is a ring road, a sample that is 90% precinct will overstate footfall quality. Stratified sampling copies the proportions, or samples each stratum enough to compare them. That is the design for “environmental quality falls with distance from the peak land-value point” if you also want to compare land-use types. Convenience sampling wrecks this: you interview outside the cafe because it is warm. State who is missing (commuters at 08:00, people who refuse). Then propose a fix: quota by age, a weekday and a Saturday, more strata. Linking the flaw to how it might have changed the conclusion is the difference between “it went well” and a level-3 evaluation.
See the idea in action
Hypothesis: pebble size decreases with distance from the cliff. Method: stratified by beach zone, then systematic every 5 m along one transect per zone, 15 pebbles at each station, longest axis. Why not convenience: the steps would oversample the lower beach. Why not pure random: tides and a groyne make some squares unsafe. Evaluation: one day’s low tide, three transects — pattern is suggestive. Improvement: more transects, random pebble selection within a quadrat at each station.
Exam technique
Turn knowledge into marks
Name the method, justify it against the question, admit convenience if you used it, and separate accuracy from reliability. “We sampled randomly” from a cluster by the coach is a gift to the examiner’s red pen.
Common mistakes
Do not give these marks away
- 01
Calling a clipboard of friends or a teacher-chosen site “random sampling”.
- 02
Naming systematic or stratified without saying why it fits the hypothesis.
- 03
Treating sample size as a vibe (“we did lots”) instead of a reliability issue.
Why is convenience sampling usually a weakness in GCSE Geography fieldwork?
ABecause it is illegal to stand near a coach
BBecause locations are chosen for ease of access rather than equal chance or fair coverage, so results can be biased and unrepresentative
CBecause it is the same as stratified sampling
DBecause random number tables are banned in schools
Show the answer
Because locations are chosen for ease of access rather than equal chance or fair coverage, so results can be biased and unrepresentative. Safety and access are understandable. The method is still biased. Top answers name the bias and propose random, systematic or stratified improvements.
Quick questions
If this is the bit you searched
What is the difference between random and systematic sampling?
Random: every unit has equal chance, using random numbers. Systematic: a regular interval along a line or list. Both beat convenience if you justify them.
When should I use stratified sampling?
When the site already has types or zones you must represent fairly — beach zones, land-use types, age groups — so one type does not swamp the sample.
Is convenience sampling ever acceptable?
It is common because of safety and time. You must label it as a limitation and show how it could have skewed results, then say how you would improve.
How big should a sample be?
Large enough that a few odd pebbles or a closed shop cannot swing the conclusion. State a number and a reliability comment; “lots” is not a size.