Use these questions to practice interrogating statistics in journalism — asking not just what the number says, but what it hides, who produced it, and what argument it is being used to support.
Governments regularly publish unemployment figures, inflation rates, and GDP growth statistics. What are some of the ways these numbers can be technically accurate but still deeply misleading? Who benefits from how these figures are calculated and presented?
Try to use: baseline, median, sample size, margin of error, data point
During a health crisis or conflict, different organizations often publish very different death tolls for the same event. How should a news consumer decide which figure to trust — or report? What does a disputed death toll reveal about the political stakes of counting?
Try to use: statistic, disputed figures, margin of error, methodology, proportion
A politician says: "Crime has risen by 40% under the opposition." A second politician says: "Crime is at its lowest level in twenty years." Both are citing official statistics. How is it possible for both to be correct — and what does this tell us about the relationship between numbers and truth?
Try to use: baseline, trend, percentage, absolute vs relative, selective presentation
Opinion polls are published constantly in the run-up to elections. What are the limitations of polling as a way of measuring public opinion? Why do polls sometimes fail dramatically — and what should we do with poll data when we encounter it?
Try to use: polling, sample size, margin of error, methodology, trend
A headline reads: "New study finds coffee DOUBLES the risk of heart disease." You read the study and find the risk rose from 0.5% to 1.0%. Is the headline accurate? Is it responsible? What obligation do journalists have when reporting on statistics from scientific studies?
Try to use: relative vs absolute, proportion, data point, methodology, margin of error
Inflation is often reported as a single national percentage figure. But inflation affects different people very differently — the poorest households typically experience much higher effective inflation than the wealthy. Why is the headline figure still reported as if it applies equally to everyone?
Try to use: average, median, proportion, baseline, trend, data point