In simple terms
A friendly intro before the formal notes — no formulas yet.
The Geographer's Detective Kit
This topic equips you with the practical tools geographers use to investigate the world. Think of it as being a detective: you gather clues (data), organise them to see patterns (presentation), and use special techniques (analysis) to solve a geographic mystery.
Imagine you're a detective investigating a series of crimes. First, you can't interview everyone in the city, so you choose a representative sample of witnesses (sampling). Then, you pin photos and notes on a large map to visualise connections (data presentation). Finally, you use forensic analysis to see if two pieces of evidence are linked (statistical tests). This process of systematic investigation is exactly what geographers do to understand spatial patterns and relationships.
- 1
Formulate a clear, testable research question or hypothesis based on a geographic theory.
- 2
Select an appropriate sampling strategy (e.g., systematic, stratified) to collect representative and unbiased field data.
- 3
Apply a suitable statistical test, such as Spearman's Rank or Chi-Squared, to analyse relationships within your data.
- 4
Interpret the statistical result in relation to your hypothesis and critical values to draw a geographically sound conclusion.
Simulations
Every simulation here runs the real model — try the steps on a card, then check what you see against the notes.
3 simulations
- GeoGebraCoreIB 3.4
Spearman's Rank Correlation Coefficient
Choose Random or Linked data, slide pattern through four shapes of relationship, and press Update Random Numbers for a fresh sample.
Try this
- Choose Linked and pattern 1: a strong positive trend gives a coefficient near +1.
- Step through the other patterns and predict the sign and size before you read the value.
- Switch to Random and press Update Random Numbers several times: note how far from 0 chance alone can push the coefficient.
Look for Spearman's Rs measures how consistently one variable rises or falls with the other. A random sample rarely gives exactly 0, which is why you compare Rs with a critical value.
Michael Borcherds · GeoGebra · GeoGebra Terms of Service
- GeoGebraCoreIB 3.4
χ² test: reject or not?
Drag the calculated χ² point along the axis and slide the degrees of freedom; the critical value, p-value and decision update.
Try this
- Leave 5 degrees of freedom and drag the calculated value up until the verdict flips to Reject H₀.
- Note the critical value at that moment (the significance level is 0.05).
- Lower the degrees of freedom and see the critical value fall.
Look for You reject the null hypothesis only when calculated χ² exceeds the critical value for your degrees of freedom; a bigger table needs a bigger χ² to be significant.
Casandra Hutchinson · GeoGebra · GeoGebra Terms of Service
- PhETIB 3.4
Least-Squares Regression
Pick a real data set or drop your own points on the graph, then tick Best-Fit Line to see the line and the correlation coefficient r.
Try this
- Choose a data set from the menu and judge the direction and strength of the relationship by eye.
- Tick Best-Fit Line and compare r with your estimate.
- Switch to Custom, plot ten points, then drag one far from the rest and watch r change.
Look for A scatter graph shows direction, strength and anomalies before any test is run, and one outlier can change a correlation a great deal. The r shown is Pearson's, not Spearman's Rs.
Simulation by PhET Interactive Simulations, University of Colorado Boulder · Licensed to MarkScheme (public licence CC BY-NC 4.0 since 2026-03-30)
Key formulas
Tap any symbol to reveal exactly what it means and its units.
Full topic notes
Formal explanation with the rigour you need for the exam.
Data Collection and Sampling Techniques
All geographical investigations begin with data. This data can be primary (collected by you) or secondary (collected by others). The quality of your conclusions depends entirely on the quality of your data collection. A crucial part of this is sampling – selecting a representative part of a population to study. The choice of sampling technique is vital for minimising bias and ensuring your results are valid.
Random Sampling: Each member of the population has an equal chance of being selected. Often done using random number generators. Pro: Unbiased. Con: Can lead to poor coverage if samples cluster by chance.
Systematic Sampling: Samples are taken at regular intervals (e.g., every 5th person, every 10 metres). Pro: Simple to implement and gives good spatial coverage. Con: Can be biased if the sampling interval matches a pattern in the population.
Stratified Sampling: The population is divided into sub-groups (strata), and a sample is taken from each. This ensures all sub-groups are represented. Pro: Highly representative. Con: Requires prior knowledge of the population to create the strata.
Statistical Analysis: Spearman's Rank Correlation
Once data is collected, we need to analyse it. Spearman's Rank is an inferential statistical test used to measure the strength and direction of a relationship between two variables. It is ideal for data that is continuous or can be ranked. The result, the correlation coefficient (), is a value between -1 and +1.
Where is the difference in rank between paired values.
And is the number of pairs of data.
A value near +1 indicates a strong positive correlation.
A value near -1 indicates a strong negative correlation.
A value near 0 indicates a weak or no correlation.
Statistical Analysis: Chi-Squared Test
The Chi-squared ($chi^2$) test is used when you have categorical data (data in named groups, like land use types or yes/no answers) and you want to see if there is a significant association between two variables. It works by comparing the frequencies you observed in your fieldwork (Observed, O) with the frequencies you would expect to get if there was no association (Expected, E).
Worked examples
See the formulas applied — reveal one step at a time, like the exam.
A student investigated the effect of a factory on air quality by measuring lichen coverage on trees at different distances. Calculate Spearman's Rank for the data below and comment on the result at the 95% significance level. (Critical value for n=8 is 0.738).
| Site | Distance from factory (m) | Lichen coverage (%) |
|---|---|---|
| 1 | 100 | 5 |
| 2 | 250 | 12 |
| 3 | 400 | 10 |
| 4 | 600 | 25 |
| 5 | 800 | 30 |
| 6 | 1000 | 45 |
| 7 | 1200 | 42 |
| 8 | 1500 | 55 |
- 1
State Hypotheses:
A geographer investigates if there is an association between beach material type and the presence of a groyne. They collected the following data. Test for an association at the 95% significance level. (Critical value for 1 degree of freedom is 3.84).
Observed Frequencies (O):
| Sandy | Shingle | |
|---|---|---|
| Groyne Present | 40 | 15 |
| No Groyne | 20 | 25 |
- 1
State Hypotheses:
How it all connects
The big idea sits in the middle — tap a linked idea to explore the link.
Tap a linked idea to see how it connects back to the main topic — that connection is what examiners reward.
Glossary
Key terms for this topic — skim now; the Check step will test them.
- Spearman's Rank Correlation Coefficient ()
A statistical test used to determine the strength and direction of a relationship between two sets of ranked, continuous data. The result ranges from -1 (perfect negative correlation) to +1 (perfect positive correlation).
- Chi-squared test ($chi^2$)
A statistical test used to determine if there is a significant association between two categorical (nominal) variables. It compares observed frequencies with expected frequencies.
- Null Hypothesis ()
A statement of no relationship or no difference between variables. Statistical tests aim to either reject or fail to reject the null hypothesis. E.g., 'There is no significant correlation between distance from the CBD and pedestrian count.'
- Significance Level (p-value)
The probability of rejecting the null hypothesis when it is actually true. A common significance level in geography is 5% (p=0.05), meaning there is a 5% chance the results are due to random chance.
- Degrees of Freedom (df)
The number of values in a final calculation of a statistic that are free to vary. For Spearman's Rank, it is the number of pairs (n). For Chi-squared, it is (rows-1) x (columns-1).
- Proportional Symbol Map
A map that uses symbols of different sizes (e.g., circles, squares) to represent the magnitude of a variable at a specific location. The area of the symbol is proportional to the value it represents.
- Choropleth Map
A thematic map where areas (e.g., countries, administrative districts) are shaded or patterned in proportion to the measurement of a statistical variable. Best used for density or ratio data, not absolute numbers.
- Isoline Map
A map that uses lines (isolines) to connect points of equal value. Examples include contour lines (elevation), isobars (pressure), and isotherms (temperature).
- Stratified Sampling
A sampling method where the population is divided into subgroups (strata) based on a shared characteristic (e.g., land use zones). A random or systematic sample is then taken from each subgroup, ensuring representation.
- Systematic Sampling
A sampling method where samples are chosen at regular intervals. For example, surveying a person every 10 metres along a transect. It is easy to implement but can be biased if the interval coincides with an underlying pattern.
- Primary Data
Data collected first-hand by the researcher for a specific purpose. Examples include questionnaire surveys, environmental quality surveys, and field measurements.
- Secondary Data
Data that has been collected by someone else and is publicly available. Examples include census data, government reports, and weather station records.
Name it
Read the meaning, then pick which of this lesson’s terms it describes. Miss one and you see what your choice really means.
A statement of no relationship or no difference between variables. Statistical tests aim to either reject or fail to reject the. E.g., 'There is no significant correlation between distance from the CBD and pedestrian count.'
Quick check
Write your answer first, then compare it with the model one — the gap is what you would have lost.
Teach it back
If you can explain it simply, you own it — gaps here are marks you’d lose.
Teach it back
Explain this topic as if teaching a friend. We name the gaps an examiner would still dock.
Revision flashcards
Guess first, then flip — retrieval beats re-reading.
Key takeaways
Review these before you close the topic — retrieval beats re-reading.
Random Sampling: Each member of the population has an equal chance of being selected. Often done using random number generators. Pro: Unbiased. Con: Can lead to poor coverage if samples cluster by chance.
Systematic Sampling: Samples are taken at regular intervals (e.g., every 5th person, every 10 metres). Pro: Simple to implement and gives good spatial coverage. Con: Can be biased if the sampling interval matches a pattern in the population.
Stratified Sampling: The population is divided into sub-groups (strata), and a sample is taken from each. This ensures all sub-groups are represented. Pro: Highly representative. Con: Requires prior knowledge of the population to create the strata.
Practice — then mark it
The whole point: a real Cambridge question, marked mark-by-mark.
Test Your Geographic Skills
Test Your Geographic Skills
Extra simulations & links
PhET, GeoGebra and other curated tools — open in a new tab.
Frequently asked
Checkpoint
One marked question is worth ten re-reads — close the loop before you move on.
Reading it isn’t knowing it — prove it.
Before you move on: do Test Your Geographic Skills on paper, snap a photo, and get examiner-style feedback on exactly where you win and lose marks.
Discuss Geographic skills — data, maps and fieldwork
Ask, share and discuss with other Geography HL students