Read a chart without missing the context
Check the title, units, axes, time period and source before drawing a conclusion from a graph.
Follow the method ↓
The answer in 30 seconds
A chart makes patterns easier to see, but its design can change the impression. Reading the labels and underlying comparison matters more than the visual shape alone.
State what is measured, in which unit and period, then inspect the axes, the comparison group and the source. Describe the pattern only as far as the data supports it.
Examples to adapt
Check the vertical axis start and scale before describing how large the change really is.
Ask whether values are totals, percentages or rates measured over the same period.
Find the original chart and its notes so the title, source and caveats are visible.
One chart, two views
Change the axis without changing the data
Fictional example: an index rises from 98 to 99 to 100 over three equally spaced periods. Compare the line on a full and a narrow scale. Its values and change stay the same.
| Period | Index, arbitrary unit |
|---|---|
| A | 98 |
| B | 99 |
| C | 100 |
From A to C: +2 units, approximately +2.04%. A line encodes positions: a clearly labelled narrow scale can reveal a change. For bars whose length encodes the value, use a zero baseline.
Documentation: Office for National Statistics — Axes and gridlines · Consulted on 15 October 2026. Display changes locally without saving.
Follow the method
- 1Read the title and source
Find who produced the chart, what was measured and whether a newer edition exists.
- 2Inspect both axes
Check units, starting values, intervals and whether time is continuous or selected.
- 3Check the comparison
Identify the populations, denominators and periods behind the plotted values.
- 4Look for missing context
Read notes on exclusions, estimates, definitions or breaks in the series.
- 5Write a careful conclusion
Describe the observed difference without turning a visual association into a cause.
Workshop: read a chart with its context
Before interpreting a line, note its unit, axis origin, time period and population. Find the underlying data or table. Then restate what changed, using absolute and relative values where possible.
A truncated axis can magnify a difference and an average can hide subgroups. Check sample sizes and method before treating the pattern as a general trend.
A closer look
Recover the numbers behind the visual impression
Fictional example: an indicator starts at 98 and ends at 100. Before commenting on a chart’s slope, identify both values and where the axis begins.
| Observed feature | What it changes | What remains true |
|---|---|---|
| Vertical axis starts at 0 | The gap looks small against the full scale | The absolute difference is +2 |
| Vertical axis starts at 97 | The gap occupies much more of the chart height | Relative change is approximately +2.04% |
| Period or population changes | The comparison may mean something different | The numbers 98 and 100 alone are insufficient |
A narrow scale can reveal small changes; read its markings carefully. Appearance alone establishes neither manipulation nor practical significance.
What sentence is supported by these values alone?
“In this series, the indicator rises from 98 to 100 between the two dates.” Explaining why requires more evidence. Comparing with another change also requires matching units, periods and populations.
A checklist to keep
Use these checks to record what you found. The grid supports a decision; it does not make one for you.
| Check | What to examine | Action |
|---|---|---|
| Title | What question does the chart address? | State |
| Axes | Units, origin and intervals | Inspect |
| Population | Who or what is counted? | Define |
| Period | Dates and possible gaps | Check |
| Source | Producer and notes | Cite |
Read the checklist directly in the table.
What to check
An axis that starts above zero can make a small difference appear larger.
Different units or denominators can make two lines incomparable.
A chart alone rarely explains why a change happened.
Common questions
Is a truncated axis always misleading?
No. It can reveal small changes, but it should be clear and appropriate to the question. Check the numeric values before judging their size.
How do I cite a chart?
Record the chart title, producer, publication date, source URL and the specific measure used.
Sources and documentation
Documentation consulted on . Examples are illustrative; interfaces and results may change.