Lesson
Data language: exact quantities, comparisons and limits
Goal: Separate counts, shares, percentage changes and percentage-point differences · Distinguish the largest level from the fastest growth · Match trend language to the observations actually supplied · Report forecasts and derived comparisons without invented causes
- Begin with the measurement, not the trend verb. A chart may show a count, a percentage of a changing total, an index or a rate per person. A rise from 20% to 30% is ten percentage points, or a 50% increase relative to the initial share. It is not a ten per cent relative rise. A larger number can coexist with a smaller share if the total grows faster. Similarly, the category with the largest absolute increase may have a smaller proportional increase because it starts from a larger base. State which comparison you mean and keep the denominator visible.
- Choose temporal language only after inspecting the available observations. Two endpoints show a net change, not necessarily steady movement between them. Four rising values can still show a slowing rate of increase. A series that rises by 30, then 20, then 10 units is still rising; its growth is slowing. The highest observed value can be called a peak within the displayed series, but a forecast should be labelled as projected and should not be reported as a completed event. Use approximate wording when reading an approximate scale; do not create false precision from a chart that does not supply it.
- Report relationships that the data support. In Academic Task 1, a change after a policy announcement does not establish that the policy caused it; even a hedged causal story is unnecessary when the visual does not provide that explanation. Compare like measures and identical periods. Derived comparisons can be useful if they are accurate and help identify a main feature, but performing every possible calculation crowds out selection. All figures in this lesson are invented for practice. Your task is to distinguish the informative, defensible statement from the fluent statement that subtly changes the unit, time span or evidential boundary.
A worked example
BeforeFictional data: Region A's share rose from 20% to 30%; its count rose from 200 to 240. Draft: 'A increased by 10%, while the number grew by 40%. This steady rise proves that the new policy worked.'
AfterRegion A's share rose by ten percentage points, from 20% to 30%, while its count increased by 40, or 20%. The figures show a rise between the two observations but do not establish a steady trend or explain its cause.
'20% to 30%' is a ten-point difference and a 50% relative increase in share. '200 to 240' is 40 divided by the starting 200, giving 20%. Two observations cannot show what happened between them, and the data do not supply a causal explanation.
Use it elsewhere
Choose an unfamiliar chart and write four statements: a level comparison, an absolute change, a proportional change and a limitation on what can be inferred. Check each calculation independently before turning them into a short report.