Did you know that a perfectly real number can become a misleading statement when the reader misses one word beside it? A report labelled advance or preliminary may be accurate for the evidence available on publication day while still being open to revision. The quiet word beside a figure changes how confidently we should repeat it.
Vocabulary matters because provisional, revised and final tell us where a statistic sits in an evidence process. Before sharing a chart, a news headline or a school project claim, check the data’s status, reference period, release date, measurement definition and source. That five-part habit helps students distinguish a useful early estimate from a later, better-informed figure.
The hidden question beside every number: how settled is it?
A statistic is a number produced using a definition and some evidence. The definition tells you what was counted; the evidence tells you how the count or estimate was obtained. The status word tells you whether the evidence is still being completed, checked or updated. Missing the status can cause an argument even when two readers quote a trustworthy source.
Imagine one news story says a school’s exhibition attracted 420 visitors while a later school report says 438. Without context, a reader might think somebody made a mistake. With context, the first number could be a provisional tally taken before late sign-in sheets were counted, and the second could be a revised total. Both could have been honestly reported at different moments. These are fictional teaching numbers, not an actual school’s results.
Six words that change the meaning of a statistic
- Estimate: a value inferred from available information rather than necessarily counted exactly. An estimate is not automatically a wild guess.
- Advance estimate: an early release made before all the more complete information is available. The agency’s own definition governs its use.
- Preliminary or provisional: an initial figure that may change after additional information, validation or reporting. Some organisations distinguish these terms, so read their notes.
- Revised: a previously issued figure has been updated. Ask what changed: new returns, corrected errors, definitions, adjustments or methodology.
- Final: the figure is designated final for a particular release or statistical process. It does not mean knowledge can never change after a later methodological review.
- Forecast or projection: a statement about a possible future outcome, usually under assumptions; it is not a measurement of what has already happened.
Related vocabulary supplies further clues. A reference period is the time the statistic describes. A publication date is when the report appeared. A benchmark is information used for comparison or adjustment. A data vintage is the particular version of a dataset available at a point in time. Seasonally adjusted and not seasonally adjusted are not interchangeable labels.
A Singapore example: what an advance GDP estimate really promises
Singapore’s Department of Statistics explains that advance quarterly GDP estimates use timely, incomplete information, with more comprehensive information incorporated in subsequent estimates. Its explanation of national accounts also describes revisions made as annual survey benchmarks and updated administrative data become available. That is a routine part of statistical production, not proof that earlier figures were dishonest.
For example, the Singapore Department of Statistics home page identified its 14 October 2026 release on third-quarter 2026 GDP as an advance estimate. A student reporting that release should say what the release was called and the date, not silently present it as the ultimate outcome for all time. The lesson is about interpreting labels; readers should consult the current report for the latest figure.
Read the primary explanations at SingStat: Our Data Explained — National Accounts and the Singapore Department of Statistics. Those sources own Singapore GDP methods; this vocabulary article does not replace them.
Worked example: three versions of a school survey
Suppose a Secondary 2 class asks students which of three transport methods they use most often. This entire example is invented to demonstrate reading status labels, not to report student behaviour. On Monday the class has received 80 forms and publishes a provisional table. On Wednesday it finds two duplicate submissions and receives additional valid forms. On Friday the teacher approves a final project table.
- Monday: provisional 48 of 80 returned forms name public transport, or 60%. The denominator is the forms received so far.
- Wednesday: revised 54 of 90 valid forms name public transport, also 60%. The numerator and denominator changed although the percentage stayed the same.
- Friday: final for the project 56 of 92 valid forms name public transport, approximately 60.9%. State the final project denominator and acknowledge the sample’s limitations.
Notice the three reasoning tasks. First, revised does not always mean the percentage increases or decreases. Second, 60% of the returned forms is not automatically 60% of the school population. Third, a final classroom tally cannot establish a national trend. The vocabulary protects the chain of claims.
The five-question evidence check
- What exactly is counted? Read the unit, population, category definitions and exclusions.
- When does it refer to? Find the reference month, quarter, term or year. Do not substitute the website’s latest update date.
- What version am I looking at? Look for provisional, advance, revised, final, correction notices and a table’s footnotes.
- What changed, and why? Compare like with like; do not mistake a revised method or sample for a real-world change.
- What may I safely claim? Quote the version, source and date; qualify uncertainty rather than pretending it has disappeared.
Word families, sentence patterns and collocations
Students often know revise as ‘study again for an exam’. In data reporting, to revise an estimate is to update it. The noun is revision; the adjective is revised. A student can therefore revise for Mathematics and still need to identify a revised percentage in a Science report. Context decides the meaning.
Try these natural combinations: preliminary findings, provisional figures, advance estimate, updated dataset, revised methodology, final count, latest release and subject to revision. Avoid saying ‘a provisional fact was wrong’ simply because a later estimate changed. More careful: ‘The preliminary estimate was later revised after additional data became available.’
How this carries into English, Mathematics and Science
Primary English and reading comprehension
In a comprehension passage, look for the verb and qualifier around the number. ‘Researchers estimated 200′ differs from ‘Officials confirmed 200′. The child should answer using the evidence level in the passage, not a stronger verb invented during retelling. Ask, ‘Which word makes the writer cautious?’
Primary Mathematics and early data handling
A bar chart can show precise-looking bars even when its source is a small survey. Students should read the title, legend, axis, unit, sample size and any note saying ‘estimate’. Calculations may be correct while a conclusion about a wider population remains unjustified. The same concern grows in secondary Mathematics, when tables are compared across years or datasets.
Secondary Science, G1/G2/G3 and later study
Lab readings may first be recorded, then checked for transcription errors or instrument issues. Distinguish a measured observation from an estimated value and a predicted outcome. When writing a Science conclusion, state the evidence actually collected. The subject-specific vocabulary may become more technical at different levels, but this habit is useful whether students study English, Mathematics or Science.
For more work on recognising what a scientific record actually contains, see eduKateSG’s guide to timestamps, sampling intervals and missing sensor data. For a broader bridge across school subjects, use How English Works: English Across School Subjects. These existing guides cover their own subjects; the present article focuses on the status of a changing number.
A ten-minute retrieval and application exercise
Give the child three mock headline cards: ‘Provisional survey estimate: 30%’, ‘Revised after late entries: 28%’ and ‘Final classroom report: 29%’. Do not reveal the definitions. Ask the child to sort them into release order, explain what additional evidence could cause a revision, and write one responsible sentence. Then change the order of the cards and ask again without the notes.
One sound answer is: ‘The classroom’s final report recorded 29%; earlier estimates were provisional and later revised.’ A stronger answer adds the class sample and report date. An unsound answer is: ‘A national survey proves that exactly 29% of all students agree.’ That conclusion was never in the example.
After two days, ask the child to recall the difference between revised and forecast without looking. After a week, transfer the skill to a weather prediction, a published research chart, or a newspaper’s economic graph. Retrieval followed by use in a new context is more revealing than rereading the glossary five times.
What parents and teachers can do tomorrow
- At home, look at one number from the news together and find its status word and date before discussing the headline.
- During reading, have the child highlight words that limit certainty: early, estimated, likely, approximately, preliminary and revised.
- At school, require each group to label a data table with source, unit, reference period and version where relevant.
- During feedback, ask students to correct a claim without changing the numerical working. This separates interpretation errors from arithmetic errors.
- Before publication or a class presentation, compare the student chart with the latest source version and record any change.
Common misunderstandings worth correcting
‘Preliminary means fake.’ No: preliminary can mean an honestly produced early estimate with a defined method. ‘Revised means unreliable.’ No: revision often reflects better or fuller evidence, though readers should inspect how and why it changed. ‘Final means permanent truth.’ No: final is a stage label within a release system. ‘Latest means most relevant.’ Not always: a later publication may describe a different period or population.
Frequently asked questions
Is a preliminary number safe to use in homework?
Yes, when it fits the question and is labelled accurately. Cite the agency, date, reference period and ‘preliminary’ status; update to the newer release if the assignment requires current data.
If two websites show different GDP numbers, which is correct?
First check whether they refer to the same quarter, comparison basis and release vintage. Advance and subsequently revised estimates can differ legitimately. Check the official source before accusing either page of an error.
Does final data always mean a complete census?
No. A final estimate may still be based on surveys or models. ‘Final’ describes a release stage, not the universal absence of estimation or sampling.
Should a Primary child learn all six terms immediately?
Start with guess, estimate, checked and confirmed, using age-appropriate examples. Add provisional, revised, final and forecasting as the child begins to read charts and news reports independently.
The takeaway
Twenty-first-century vocabulary is not only a list of impressive words. It tells a reader how much a claim can carry. A careful child who sees ‘revised’ beside a figure asks better questions, writes more accurately, and is less likely to pass on a number without its meaning. Keep the definition, source, date and version attached to the number—and the number becomes genuinely useful.