Official counts often change after first publication because records arrive late, errors are corrected, or classifications improve.
Common reasons figures move
Late records and new classifications
Find the agency's revision policy and compare preliminary with final figures. Some series are updated on a fixed schedule; others change when better records become available. Revisions can be small or material, so a responsible account states which release it uses. A revised number does not automatically mean manipulation, just as an official label does not make a figure immune to error.
The first release is a draft view
An early monthly count misses forms submitted near the deadline. The next release adds them, making the prior month look different without altering what happened on the ground.
Keep the comparison consistent
When quoting a changing series, record the release date and whether the figure is preliminary. Link to the table rather than only a screenshot.
What a revision note reveals
Locate two releases of the same series. Note whether the previous period changed and whether the explanation identifies late reporting, reclassification, or another reason.
Original editorial illustration
Cite the version you used
Avoid comparing figures from different vintages without saying so. Check that categories and time periods still match.
Follow the evidence trail: Why administrative data get revised
Return to the question when fresh evidence appears. A revised dataset, a follow-up study, a final decision, or a change in method can alter the answer without making the original inquiry pointless. Record what changed and whether it affects the main conclusion or only a detail. A useful correction is visible and specific: it lets a reader understand the earlier statement, the new evidence, and the reason for the updated wording.
A short explanation often hides several decisions about which source to use, which period to describe, and which people or places to include. Save the original record alongside the summary and note the date of each. If a later account changes a definition or introduces a correction, the earlier version remains useful for seeing how the interpretation evolved. A repeated claim is not independent confirmation when every repetition points back to the same unexamined source.
What could change the answer
Before drawing a practical conclusion, write down the comparison that would answer the real question. A national average may conceal a local difference; a single striking case may be unrepresentative; and a proposal may have no effect until it is adopted and implemented. Keep the unit, denominator, time window, and any missing observations beside the number. This makes it easier to tell whether two apparently similar figures can be compared at all.
The first release is a draft view
Late records and new classifications
Cite the version you used
It helps to separate three lines in your notes: what the source directly records, what you infer from it, and what remains open. The first line should be traceable to a document or observation. The second should explain the reasoning and alternatives. The third is where missing data, possible revisions, and competing explanations belong. If the second line becomes more confident than the first can support, narrow the wording rather than hiding the uncertainty.