Features
Part of Magazine copy editing: a practical guide to clean, accurate copy
Case study: fact-checking a fictional magazine data claim
Fictional fact-check case where a magazine corrects a survey claim, percentage-point error, weak comparison, overstated headline, caption date, and source record.
What to take away
- A dramatic percentage in the draft used the wrong denominator.
- The source table reported estimates with uncertainty, not exact population counts.
- A year-to-year comparison used differently worded survey questions.
- The headline and chart repeated a stronger claim than the body.
- The final article described the supported finding and preserved the correction trail.
This fictional case shows an editorial process. The magazine, survey, people, figures, and quotations are invented. The example does not describe a real study or publication.
The submitted claim
The fictional magazine Local Measure receives a feature about evening bus use. The draft says, "Late-shift ridership rose 40 percent in one year, proving that a pilot route transformed commuting." A chart shows 20 percent for 2025 and 28 percent for 2026. The headline calls the result "a 40-point leap."
The fact-checker separates the sentence into four claims: the measured population, the size of change, the time comparison, and causation. The source packet contains two survey summaries, a transit-agency press release, a spreadsheet exported from an online dashboard, and an interview with the pilot manager.
Step 1: recover the original data
The checker requests the questionnaires, methods, full tables, revision dates, and definitions. The 2025 survey asked whether respondents used any bus after 8 p.m. during the previous month. The 2026 survey asked whether respondents used the new route after 7 p.m. during the previous week. The percentages do not measure the same behavior.
The International Fact-Checking Network's Code of Principles calls for transparent source selection, reliance on the best available primary evidence, enough detail for readers to understand the checking method, and an open corrections policy. Guided by those principles, the team records why the dashboard alone cannot support the comparison.
Step 2: correct the arithmetic language
Even if the measures were comparable, moving from 20 percent to 28 percent is an increase of 8 percentage points and a relative increase of 40 percent. The headline's "40-point" phrase is wrong. The checker adds both calculations to the claim sheet and marks the original chart label for replacement.
Step 3: test uncertainty and causation
The tables label the figures as sample estimates and provide margins of error. They do not establish a precise change in the full rider population. The pilot also began alongside a fare discount and two large employers changing shift schedules. The available material cannot isolate the route as the cause.
The Journalist's Resource's tip sheet on the margin of error explains that data from a sample never represents the whole population exactly, that the margin of error measures how precise the estimate is and shrinks as the sample grows, and that a lead of two points inside a margin of error of three should be reported as too close to tell rather than as a lead. The fictional team asks those questions before publication rather than after a reader does.
Step 4: inspect the whole package
The checker finds that a bus-stop caption says the photograph was made on opening day, but the image metadata places it five days later. The caption becomes "during the pilot's first week." A pull quote attributed to the manager is faithful, but the layout drops a phrase that limited the claim to surveyed riders. The full phrase is restored.
The final wording
The published feature does not calculate a trend from the two surveys. It reports that each survey found use of late service among a defined sample, explains that the questions and periods differ, and presents agency records on route boardings as a separate measure. It describes the pilot, discount, and employment changes without assigning isolated causation.
The headline becomes "What a New Evening Bus Survey Can and Cannot Show." The chart displays each question verbatim beside its estimate and uncertainty. The source note explains the noncomparable periods.
Decision record
| Item | Initial wording | Final decision |
|---|---|---|
| Trend | 40 percent rise | Removed as noncomparable |
| Headline | 40-point leap | Replaced |
| Cause | Route transformed commuting | Narrowed to observed measures |
| Caption | Opening day | First week |
| Chart | Bare percentages | Questions, periods, estimates, uncertainty |
Common questions
Why not average the surveys?
Their questions, reference periods, and samples are not established as comparable.
Is 20 percent to 28 percent a 40 percent increase?
Arithmetically, yes, but only when the underlying measures are comparable and suitably precise.
Does an agency press release count as primary evidence?
It is primary evidence of the agency's statement, not automatically of the underlying result.
Was a public correction required?
No, because the errors were caught before release. The internal decision record was still preserved.
