Read beyond the noise.

Provisional Figures Need a Date and a Route to the Revision

A journalist comparing two dated statistical printouts at a desk beside a laptop.
AI-generated photographic illustration; not documentary photography.

A news story may publish a figure before every response, record, or late submission has arrived. The word “provisional” explains something about that process. It does not automatically mean the figure is careless, and it does not mean a later revision should disappear from the reader’s view.

Start with the release itself. Find the period being measured, the publication date, and the producer’s explanation of revisions. Those dates answer different questions: a report released this week may describe activity several months earlier.

Compare figures from a consistent version

Imagine a fictional series that first reports 120 events in January and later revises January to 135. If February records 140, comparing it with the unrevised January figure produces a different change from comparing it with the updated one. The arithmetic can be correct while the choice of versions misleads.

When a chart combines months, check whether earlier entries have been refreshed from the same release. A saved screenshot from last month is useful as a historical record of what was reported then, but it may no longer be the current comparison.

Find out what is allowed to change

A producer might revise for late information, improved estimates, or a changed method. Those explanations should not be treated as interchangeable. Read the accompanying notes and avoid guessing a motive from the size or direction of a revision alone.

  • Save the source link and release date with the figure.
  • Keep the label “provisional” when quoting it.
  • Check whether the producer publishes a revision calendar or history.
  • State which release supplied comparisons used in a chart.

For a reader, the useful question is not simply whether a number ever changed. Ask whether the publication made its status clear and whether the newer evidence changes the conclusion being drawn. A headline about a small increase may need more caution when that increase is smaller than the revisions routinely affecting the series.