You can easily spot a trend in a clear picture, but try finding one hidden in a massive spreadsheet of 500 numbers. Graphs translate raw data into visual stories, making geographical patterns instantly clear. Choosing the right graph and plotting it accurately are core geographical skills that turn confusing numbers into useful evidence.
Before drawing any graph, you must identify the type of data you are working with.
Categorical data (or discrete data) falls into distinct groups or counts of whole numbers, such as types of land use or the number of shops.
Continuous data is numerical data that can take any value on a scale with no gaps, such as a temperature of 15.6°C or a river's discharge.
Bivariate data consists of two sets of data used to identify a relationship or correlation, such as comparing GNI per capita against Life Expectancy.
When plotting graphs, you must identify the correct axes for your variables.
The independent variable is the factor that is changed or selected by the researcher (the "cause"). In geography, this is often time or distance, and it is always plotted on the horizontal X-axis.
The dependent variable is the factor being measured (the "effect"), which relies on the independent variable. It is always plotted on the vertical Y-axis.
To identify variables, use this mnemonic sentence: "[Independent Variable] causes a change in [Dependent Variable], and it is impossible that [Dependent Variable] could cause a change in [Independent Variable]." For example, time causes a change in traffic volume; traffic volume cannot cause a change in time. Therefore, time goes on the X-axis.
To suggest an appropriate graph, you must match the graph to the data type. A common rule is to use bar charts for categorical data and line graphs or histograms for continuous data. Do NOT draw a line graph for categorical data (e.g., connecting "Site A" and "Site B"); this is a frequent error because it implies a transition between categories that does not actually exist.
When setting up a graph, you must create a linear scale, where equal distances represent equal changes in value. If a scale is not provided, calculate the available space to ensure the graph fills at least 75% of the grid. Calculate the required interval: , then divide this by the number of squares.
When constructing your graph, you must follow a step-by-step process to ensure accurate axis labelling:
Step 1: Add a clear title to the X-axis identifying the independent variable.
Step 2: Include the correct units of measurement in brackets for the X-axis (e.g., 'Distance (m)').
Step 3: Repeat this process for the Y-axis, labeling the dependent variable and its units (e.g., 'Temperature (°C)').
When axes and scales are already provided, plotting data points accurately requires a systematic step-by-step approach:
Step 1: Find the correct coordinate value for the independent variable on the horizontal X-axis.
Step 2: Move vertically up the grid from that X-value until you align exactly with the corresponding dependent variable value on the vertical Y-axis.
Step 3: Verify your alignment with both the X and Y axes using a clear ruler if necessary.
Step 4: Mark the point with a neat, sharp 'X' rather than a dot. The center of the 'X' must be within a tolerance of half a small square (approximately 1mm) of the exact value to be awarded the mark.
Always check for broken axes, which start at a higher value rather than zero; misplotting (0,0) in the corner is a very common error.
Bar charts are used specifically for categorical or discrete data. To construct a bar chart, you must draw bars of equal width. Crucially, the bars must be separated by equal gaps (they do not touch) to show that the categories are distinct.
Histograms are used for continuous data grouped into equal intervals, known as bins (e.g., pebble lengths of 0–10mm, 10–20mm). Because the data flows continuously, the bars must touch with absolutely no gaps. To construct a histogram, draw the X-axis as a continuous scale and plot bars so the next interval starts exactly where the previous one ends. A population pyramid is a specialized back-to-back histogram showing the age and sex structure of a population.
Line graphs show changes or trends over time or distance. Compound line graphs show multiple layers of data simultaneously. In a compound graph, the value for a specific category is the vertical difference between two lines, not the height from the X-axis to the top line.
When joining points on a standard line graph, use a ruler to draw straight lines between points. However, there is a major exception: when drawing a river cross-section or long profile, you must use a smooth curve rather than ruled straight lines to accurately represent the natural shape of the landform.
Pie charts are divided circles used to show proportions or "parts of a whole" for categorical data. The entire circle represents 360°, and the data must total 100% or the total frequency. Proportional circles are a related technique where the area of the entire circle represents a total value (like total population), allowing comparison between different map locations.
To construct a pie chart from raw data, use the formula:
Step 1: Calculate the angle for each category.
Step 2: Draw a line from the center to the 12 o'clock position to start.
Step 3: Measure the largest angle clockwise and draw the first segment.
Step 4: Measure the second angle from the end of the previous segment (not from 0°).
Step 5: Continue plotting in descending order of size and add a key.
Scattergraphs show the relationship between two numerical variables. To construct one, plot the X and Y coordinates with an 'X', but do NOT connect the dots. Once plotted, draw a single straight line of best fit with a ruler, ensuring roughly an equal number of points lie on either side of the line.
A positive correlation occurs when Y increases as X increases (the line slopes bottom-left to top-right). A negative correlation means as X increases, Y decreases. You can use the line of best fit for interpolation (estimating a value within the plotted data range, which is reliable) or extrapolation (predicting a value outside the data range, which is less reliable). Points that lie far from the general trend are residuals or anomalies.
A cumulative frequency curve shows the "running total" of frequency across a dataset. It is plotted as an S-shaped curve (an Ogive) and is highly useful for comparing the sorting of sediment in Paper 3 fieldwork.
Step 1: Create a cumulative frequency column by adding each frequency to the sum of the previous ones.
Step 2: Plot points using the Upper Class Boundary for the X-axis and the cumulative frequency for the Y-axis.
Step 3: Join the points with a smooth, continuous curve.
This graph allows you to find key values: the median (at 50% of the data), the lower quartile (at 25%), and the upper quartile (at 75%). The interquartile range is the spread of the middle 50% of the data, calculated as the upper quartile minus the lower quartile.
Geography uses several specialized graphs for analyzing spatial data. Triangular graphs show the relationship between exactly three variables that must total 100% (e.g., sand, silt, and clay in a soil sample). To read them, you must follow grid lines that run at a 60° angle.
Dispersion graphs plot a single set of data along a vertical axis to show the range and spread of values at a single site, making it very easy to identify the median and anomalies. Finally, located graphs are standard charts (like pie or bar charts) placed directly onto a map to show how data varies across different spatial locations.
Students often plot cumulative frequency against the midpoint of the class interval. You must always plot it against the UPPER class boundary.
In 'Suggest' questions asking for a graph type, you must name the graph AND justify it using the data type (e.g., 'A bar chart is appropriate because the land use data is categorical').
Examiners use a strict half-square (1mm) tolerance rule for plotting. Always use a sharp HB pencil and mark points with a neat 'X' rather than a large blob to ensure you do not lose accuracy marks.
Never force your line of best fit through the origin (0,0) on a scattergraph unless the data explicitly supports it, and do NOT connect the dots.
When constructing a pie chart in an exam, always measure the next angle from the end line of the previous segment, not from 0°, to avoid losing marks to cumulative measuring errors.
A major distinction examiners look for is whether bars touch. Ensure you leave gaps in a bar chart, but ensure there are no gaps when drawing a histogram.
Categorical data
Data that falls into distinct, non-overlapping groups or categories, such as land use types.
Discrete data
Numerical data that can only take specific whole number values, such as the number of pedestrians.
Continuous data
Numerical data that can take any possible value on a scale with no gaps, such as temperature or river depth.
Bivariate data
Sets of paired data used to compare two different variables and identify if a correlation exists.
Independent variable
The variable that is changed or controlled in a study (the cause), always plotted on the X-axis.
Dependent variable
The variable being measured or tested (the effect), which changes in response to the independent variable and is plotted on the Y-axis.
Linear scale
A graph scale where equal distances along the axis represent equal changes in value.
Coordinate
A pair of values used to locate a specific point on a graph, consisting of an X-value and a Y-value.
Broken axes
An axis that starts at a higher value rather than zero, often used to skip empty space on a graph.
Bar chart
A graph for categorical or discrete data where bars are of equal width and separated by equidistant gaps.
Histogram
A graph for grouped continuous data where bars touch with no gaps to represent the continuous flow of values.
Population pyramid
A specialized back-to-back histogram showing the age and sex structure of a demographic population.
Compound line graph
A line graph showing multiple layers of data, where the value of a category is the vertical difference between two lines.
Pie chart
A circular graph divided into sectors representing numerical proportions, where the total circle equals 360° or 100%.
Proportional circle
A map technique where the total area of a circle represents a total value, allowing spatial comparisons between locations.
Scattergraph
A graph plotting bivariate data as individual unconnected points to identify correlations.
Line of best fit
A single straight line drawn through the center of a scattergraph trend, with roughly equal points on either side.
Positive correlation
A relationship where both variables increase together.
Negative correlation
A relationship where one variable increases as the other decreases.
Interpolation
Estimating an unknown value that falls within the range of plotted data points.
Extrapolation
Predicting a value that falls outside the range of plotted data points, which is generally less reliable.
Residuals
Data points on a scattergraph that lie far away from the line of best fit, also known as anomalies.
Cumulative frequency
A running total of frequencies across a dataset, plotted as an S-shaped curve.
Median
The middle value in a dataset, found at 50% of the total cumulative frequency.
Lower quartile
The value at 25% of the total cumulative frequency.
Upper quartile
The value at 75% of the total cumulative frequency.
Interquartile range
The mathematical spread of the middle 50% of the data, calculated by subtracting the lower quartile from the upper quartile.
Triangular graph
A graph with three axes arranged in an equilateral triangle, used to plot three variables that sum exactly to 100%.
Dispersion graph
A graph that plots individual data points along a single vertical axis to easily show the range, spread, and median of a dataset.
Located graph
A graph, such as a pie or bar chart, drawn directly onto a map to represent data at its specific geographical location.
Put your knowledge into practice — try past paper questions for Geography
Categorical data
Data that falls into distinct, non-overlapping groups or categories, such as land use types.
Discrete data
Numerical data that can only take specific whole number values, such as the number of pedestrians.
Continuous data
Numerical data that can take any possible value on a scale with no gaps, such as temperature or river depth.
Bivariate data
Sets of paired data used to compare two different variables and identify if a correlation exists.
Independent variable
The variable that is changed or controlled in a study (the cause), always plotted on the X-axis.
Dependent variable
The variable being measured or tested (the effect), which changes in response to the independent variable and is plotted on the Y-axis.
Linear scale
A graph scale where equal distances along the axis represent equal changes in value.
Coordinate
A pair of values used to locate a specific point on a graph, consisting of an X-value and a Y-value.
Broken axes
An axis that starts at a higher value rather than zero, often used to skip empty space on a graph.
Bar chart
A graph for categorical or discrete data where bars are of equal width and separated by equidistant gaps.
Histogram
A graph for grouped continuous data where bars touch with no gaps to represent the continuous flow of values.
Population pyramid
A specialized back-to-back histogram showing the age and sex structure of a demographic population.
Compound line graph
A line graph showing multiple layers of data, where the value of a category is the vertical difference between two lines.
Pie chart
A circular graph divided into sectors representing numerical proportions, where the total circle equals 360° or 100%.
Proportional circle
A map technique where the total area of a circle represents a total value, allowing spatial comparisons between locations.
Scattergraph
A graph plotting bivariate data as individual unconnected points to identify correlations.
Line of best fit
A single straight line drawn through the center of a scattergraph trend, with roughly equal points on either side.
Positive correlation
A relationship where both variables increase together.
Negative correlation
A relationship where one variable increases as the other decreases.
Interpolation
Estimating an unknown value that falls within the range of plotted data points.
Extrapolation
Predicting a value that falls outside the range of plotted data points, which is generally less reliable.
Residuals
Data points on a scattergraph that lie far away from the line of best fit, also known as anomalies.
Cumulative frequency
A running total of frequencies across a dataset, plotted as an S-shaped curve.
Median
The middle value in a dataset, found at 50% of the total cumulative frequency.
Lower quartile
The value at 25% of the total cumulative frequency.
Upper quartile
The value at 75% of the total cumulative frequency.
Interquartile range
The mathematical spread of the middle 50% of the data, calculated by subtracting the lower quartile from the upper quartile.
Triangular graph
A graph with three axes arranged in an equilateral triangle, used to plot three variables that sum exactly to 100%.
Dispersion graph
A graph that plots individual data points along a single vertical axis to easily show the range, spread, and median of a dataset.
Located graph
A graph, such as a pie or bar chart, drawn directly onto a map to represent data at its specific geographical location.