Types of a Civilisation | Official Statistics, National Data Quality and Public Evidence Systems

Quick answer: An official-statistics civilisation develops independent, capable institutions for producing trustworthy information about the economy, society and environment. It combines surveys, administrative records, quality standards, transparent methods, privacy, fair access, responsible revision and cooperation. A census count is not a sample survey, a headline mean is not every person’s experience, and an impressive chart cannot become sound public evidence without its numerator, denominator, source, reference period and limitations.

Did you know? A number can be calculated correctly and still mislead readers if it came from the wrong sample, used an outdated definition or hid nonresponses. The United Nations’ ten Fundamental Principles of Official Statistics were formally endorsed by the General Assembly in 2014. They emphasise trustworthy public information, professional methods, confidentiality, transparency and international cooperation—not political marketing or the public release of private household files.

Our imaginary statistical society is Dataville. Wintour asks how reports affect ordinary people and whether records are inclusive and private; Rainbolt changes one source or assumption and requires a visible correction. Every respondent, dataset, percentage and reported household in the learning exercises is fictional. They are not observations about Singapore or another real country.

This guide extends Types of a Civilisation without duplicating Demographic Change and Intergenerational Planning Networks or Land Surveying, Cadastres and Public Mapping Institutions. Those chapters explore demographic and location-specific information; this chapter focuses on the institutions and professional methods that produce official statistical evidence across fields.

The article is statistical literacy, not personal data collection, research ethics authorisation, an officially recognised survey design, an institutional data protection review or a live government indicator. All practical activities use printed imaginary numbers. Sensitive personal information should remain protected under applicable laws and institutional responsibilities.

How to use this page: five reading routes

Parents: read the opening and Missions 03, 06, 07, 11 and 16.
Primary students: separate population counts from surveys, missing values and invented examples.
Secondary learners: calculate means, medians, proportions, response rates and source-limited inferences.
Tutors: pair two of the sixteen missions each week and rotate source, mathematics and respondent-rights roles.
Researchers: start with UN Fundamental Principles, the 2019 NQAF manual, the 2025/A UNSD handbook and Singapore’s statistical-system description.

Six capacities of a trustworthy statistical civilisation

An institutional mandate for neutral, useful and accessible public statistics.
Qualified collection, classification and analysis using suitable data sources.
Measured quality, transparent metadata and clearly identified revisions.
Legal and practical protection for confidential individual records.
Respectful user engagement, correction of misuse and understandable releases.
Continuity and cooperation within and across national statistical systems.

Fifteen primary institutional reference files

UN Fundamental Principles of Official Statistics
UNECE: Ten Fundamental Principles Explained
UNSD: Management and Organization of National Statistical Systems Handbook
UNSD: National Quality Assurance Framework Manual
UNSD: NQAF Expert Group
Singapore Department of Statistics: Singapore Statistical System
Singapore DOS: Statistical Best Practices 2020
UNSD: Global Statistical Geospatial Framework
UNSD: Fundamental Principle 6 — Confidentiality
United Nations Sustainable Development Goals Data Portal
United Nations SDG Indicator Metadata
UNSD: Principles of International Statistical Activities
eduKateSG: Demographic Change and Intergenerational Planning Networks
eduKateSG: Land Surveying, Cadastres and Public Mapping Institutions
eduKateSG: Knowledge-Centred and Scientific Institutions

Official source file 1: UN Fundamental Principles of Official Statistics

The ten principles were endorsed by the United Nations General Assembly in 2014, emphasising impartiality, professional methods, transparency, confidentiality and cooperation.

Read the originating source. Identify the legal or professional role of the issuer, the edition or publication year and whether the document sets principles, methods or actual observed results. A UN statistical framework may guide national institutions but does not certify the quality of Dataville’s invented dataset. Singapore’s statutory arrangements apply in their own national jurisdiction.

The class assigns source historian, numerical auditor and public-interest reviewer. The source historian identifies precisely what can be supported, the auditor records unit, denominator and date, and the reviewer asks whose data or privacy are at stake. The learners write one credible conclusion and a second sentence that names a remaining limit.

Official source file 2: UNECE: Ten Fundamental Principles Explained

UNECE’s accessible explanation sets out ten principles including public relevance, professional independence, accountable methods and protection of individual records.

Official source file 3: UNSD: Management and Organization of National Statistical Systems Handbook

The fourth-edition handbook was endorsed by the UN Statistical Commission in 2021 and its 2025/A update adds guidance on national data ecosystems and governance.

Official source file 4: UNSD: National Quality Assurance Framework Manual

The UN Statistical Commission adopted the NQAF manual and recommendations in 2019; it explains quality frameworks, sources, measurement and responsibilities.

Official source file 5: UNSD: NQAF Expert Group

The UN expert group supports quality-assurance development, evaluation and implementation across national statistical systems.

Official source file 6: Singapore Department of Statistics: Singapore Statistical System

Singapore uses a decentralised statistical system led by DOS, working with gazetted Research and Statistics Units under the Statistics Act and relevant confidentiality requirements.

Official source file 7: Singapore DOS: Statistical Best Practices 2020

The 2020 handbook describes statistical collection, processing, quality and responsible dissemination, drawing on international statistical standards.

Official source file 8: UNSD: Global Statistical Geospatial Framework

The five framework principles connect authoritative statistical records and geography in ways that are consistent, interoperable and useful without treating every location-linked record as public.

Official source file 9: UNSD: Fundamental Principle 6 — Confidentiality

Individual data for statistical purposes are to be protected and used for appropriate statistical aims; an aggregate report and identified household file have different access rules.

Official source file 10: United Nations Sustainable Development Goals Data Portal

The SDG data portal provides indicators, sources and metadata for defined reference years; missing information and differing national reporting methods require careful interpretation.

Official source file 11: United Nations SDG Indicator Metadata

SDG metadata specify definitions, concepts, computation, data sources, limits and custodians for indicators, helping readers avoid cross-indicator comparisons without context.

Official source file 12: UNSD: Principles of International Statistical Activities

These principles promote relevance, impartiality, professional methods, confidential data and responsible treatment of erroneous statistical interpretations.

Official source file 13: eduKateSG: Demographic Change and Intergenerational Planning Networks

The companion analyses age, births and population change; this article deals with the national statistical institutions underpinning multiple kinds of evidence.

Official source file 14: eduKateSG: Land Surveying, Cadastres and Public Mapping Institutions

The companion focuses geospatial records and land relationships; this chapter addresses statistical methods and quality across topics.

Official source file 15: eduKateSG: Knowledge-Centred and Scientific Institutions

The companion examines research broadly, whereas official statistics is a distinct public evidence system with formal legal and professional obligations.

Sixteen Dataville missions

01 — Official statistics are public evidence, not government marketing
02 — National statistical office and specialised public data units
03 — Census, sample survey and administrative register
04 — Question wording, definitions and measurement validity
05 — Frames, sampling weights and who the sample can represent
06 — Nonresponse, missing values and misleading zeros
07 — Statistical disclosure control and private records
08 — Impartial methods and professional independence
09 — Comparable classifications across countries and years
10 — Revisions, preliminary estimates and statistical truthfulness
11 — Mean, median, outliers and distributions
12 — Population versus geographic denominators
13 — Quality assurance across the complete statistical workflow
14 — SDG indicators, targets and missing data
15 — AI, automated data pipelines and independent checking
16 — Statistical archives and trust beyond the founding team

Mission 01 — Official statistics are public evidence, not government marketing

The Dataville example: In fictional Dataville, an agency publishes impressive charts selected only to support its director’s favourite policy.

The statistical question and immediate useful answer

The UN Fundamental Principles state that official statistics should serve public information needs and be produced and made available impartially on professional grounds. A statistic’s provenance, method and relevance matter independently of whether it supports a particular policy.

Start with a decision-maker’s real information question, such as how many households need a service or whether a particular school programme changed over time. Next define the population, reference period, statistical concept and acceptable source. Avoid letting a decorative dashboard determine the policy question after the fact.

Here the initial public assertion is: In fictional Dataville, an agency publishes impressive charts selected only to support its director’s favourite policy. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: The UN Fundamental Principles state that official statistics should serve public information needs and be produced and made available impartially on professional grounds. A statistic’s provenance, method and relevance matter independently of whether it supports a particular policy. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

Different sources may be useful for different aims. A register collects information for administrative purposes, a survey can estimate selected characteristics with a sample, and a census aims at broad enumeration under its defined frame. Every approach has coverage, cost, privacy and timeliness considerations. Describe them rather than insisting the largest dataset is always best.

The statistical workflow is an institution, not a spreadsheet alone. Design, collect, validate, link, process, calculate, review, publish, archive and correct are distinct tasks with suitable staff and controls. A result may be arithmetically sound but unfit for use because the data collection frame omitted relevant people.

For official statistics are public evidence, not government marketing, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Impartiality does not mean every interpretation is equally supported. A statistical office should choose professionally grounded methods and disclose the evidence, including inconvenient results and uncertainty. It should be able to comment on misuse of its figures without telling every reader which political policy to prefer.

Public trust depends on more than one impressive figure. Statistical releases need source notes, accessible formats, methodology, honest limitations and a way to contact the responsible producer. If a graph’s design conceals the meaning of its axis, the caption has not done its job even when the plotted points are numerically correct.

Check whether the appropriate primary source concerns the Fundamental Principles, Singapore’s statistical system, the UN NQAF or SDG indicator metadata. An international principle cannot supply local private records. A national office’s published series should be used with its actual reference dates and definitions. Good public information can explain scientific uncertainty without endorsing a policy party.

Work through the mathematics, then examine the denominator

In a toy report ten indicators exist and four appear in a selective publicity card. Forty percent shown is a reporting choice, not evidence the omitted indicators are false.

The numerical example is an exercise, not a measured public fact. State the numerator, denominator, units and observation year, and keep a zero separate from missing or suppressed data. The student should explain whether a mean, median, rate, sample proportion or administrative count is actually relevant.

Change one imaginary input and calculate again. Ask whether an average hides a distribution, whether a response rate is representative, or whether area and population are being confused. A revised percentage should not automatically be called improvement in welfare, safety or government effectiveness.

Rainbolt reveals the changed evidence

Rainbolt discovers an omitted source contradicts the headline’s sweeping claim.

Rainbolt changes a fundamental element: a source definition, a revised data vintage, duplicate records, an inaccessible survey question, a different weighting scheme or a new privacy concern. Preserve the old claim, calculate or analyse the effect, and issue a dated correction. Altered evidence should be visible rather than quietly buried.

Keep both versions of the statistic in a visible audit trail. Explain which assumption or transformation changed and which earlier interpretation has become unsupported. A versioned correction is not a sign that statistical reasoning is futile; it is part of scientific integrity and public accountability.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks whether all readers can access appropriate information regardless of their views.

Privacy is a central professional responsibility. Records about people and businesses can include sensitive attributes; a public chart can reveal a group trend without publishing a household’s individual information. Conversely, a tiny cell or joined datasets can still reveal individuals. Use entirely fictional records to teach this principle.

Wintour asks who is missing from the report and who bears consequences when the statistical picture is wrong. People lacking Internet, groups omitted from administrative systems, smaller firms and people with disabilities may be underrepresented. Inclusion requires competent design, confidentiality and attention to the needs of users.

Ask whether a dataset might exclude a group or reveal private information through a cell, join or careless caption. The class should protect people’s agency and refrain from treating a statistical category as a personal identity. Legitimate data governance balances public knowledge with rights and acceptable research purposes.

Safe three-person tutorial and family activity

Compare invented press claims against a complete source-labelled table.

International comparability is an achievement requiring sustained cooperation and harmonised classifications, not an automatic consequence of publication in English. Singapore’s statistical system has national legal and professional structures; UN and OECD frameworks may inform questions without imposing identical definitions or rules in every jurisdiction.

Three learners rotate a source-and-quality reader, a computation and denominator checker and a respondent-and-user advocate. After a correction each explains a concern raised by another role. This helps avoid both shallow numerical confidence and a vague claim that no evidence can ever be trusted.

The final document should state data owner, purpose, population, source, collection period, statistic definition, method, missing-data treatment, privacy status, any revision and a question remaining for professional review. When a new officer or researcher takes over, the statistical result should still be intelligible.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: In fictional Dataville, an agency publishes impressive charts selected only to support its director’s favourite policy. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: The UN Fundamental Principles state that official statistics should serve public information needs and be produced and made available impartially on professional grounds. A statistic’s provenance, method and relevance matter independently of whether it supports a particular policy. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For official statistics are public evidence, not government marketing, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: In a toy report ten indicators exist and four appear in a selective publicity card. Forty percent shown is a reporting choice, not evidence the omitted indicators are false. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt discovers an omitted source contradicts the headline’s sweeping claim. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks whether all readers can access appropriate information regardless of their views. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain official statistics are public evidence, not government marketing, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 02 — National statistical office and specialised public data units

The Dataville example: Dataville assumes every government agency makes figures under identical definitions and the national office controls every underlying case record.

The statistical question and immediate useful answer

Statistical systems may be centralised or decentralised. Singapore DOS coordinates a national system with specialised Research and Statistics Units, subject to the Statistics Act. Institutional responsibilities, legal mandates and statistical definitions differ by topic.

Here the initial public assertion is: Dataville assumes every government agency makes figures under identical definitions and the national office controls every underlying case record. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: Statistical systems may be centralised or decentralised. Singapore DOS coordinates a national system with specialised Research and Statistics Units, subject to the Statistics Act. Institutional responsibilities, legal mandates and statistical definitions differ by topic. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For national statistical office and specialised public data units, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

In a fictional diagram a national office coordinates six specialist units but four have a documented data-sharing standard, a two-thirds documentation count rather than data quality.

Rainbolt reveals the changed evidence

Rainbolt reveals two units label employment using incompatible age groups.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks how citizens know who maintains and corrects each public series.

Safe three-person tutorial and family activity

Draw a paper producer, coordinator, method-owner and public-user map.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville assumes every government agency makes figures under identical definitions and the national office controls every underlying case record. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: Statistical systems may be centralised or decentralised. Singapore DOS coordinates a national system with specialised Research and Statistics Units, subject to the Statistics Act. Institutional responsibilities, legal mandates and statistical definitions differ by topic. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For national statistical office and specialised public data units, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: In a fictional diagram a national office coordinates six specialist units but four have a documented data-sharing standard, a two-thirds documentation count rather than data quality. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt reveals two units label employment using incompatible age groups. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks how citizens know who maintains and corrects each public series. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain national statistical office and specialised public data units, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 03 — Census, sample survey and administrative register

The Dataville example: A Dataville chart treats responses from twenty volunteer families as a perfect count of every household.

The statistical question and immediate useful answer

Census-like enumeration, statistical surveys and administrative records have different coverage, costs, reporting delays and selection biases. The national statistical institution must evaluate each source for its intended quantity, not assume the largest dataset is always the best.

Here the initial public assertion is: A Dataville chart treats responses from twenty volunteer families as a perfect count of every household. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: Census-like enumeration, statistical surveys and administrative records have different coverage, costs, reporting delays and selection biases. The national statistical institution must evaluate each source for its intended quantity, not assume the largest dataset is always the best. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For census, sample survey and administrative register, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

A toy community has 200 households and an online survey reaches 50. Twenty-five percent response coverage is not proof of a representative sample.

Rainbolt reveals the changed evidence

Rainbolt finds the survey excludes homes without broadband.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks about fair representation and the privacy costs of unnecessary questions.

Safe three-person tutorial and family activity

Sort paper source cards by frame, selection method, possible omission and intended statistic.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: A Dataville chart treats responses from twenty volunteer families as a perfect count of every household. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: Census-like enumeration, statistical surveys and administrative records have different coverage, costs, reporting delays and selection biases. The national statistical institution must evaluate each source for its intended quantity, not assume the largest dataset is always the best. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For census, sample survey and administrative register, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: A toy community has 200 households and an online survey reaches 50. Twenty-five percent response coverage is not proof of a representative sample. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt finds the survey excludes homes without broadband. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks about fair representation and the privacy costs of unnecessary questions. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain census, sample survey and administrative register, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 04 — Question wording, definitions and measurement validity

The Dataville example: Dataville asks residents whether they are ‘satisfied and successful’ using one question and reports the percentage as measured wellbeing.

The statistical question and immediate useful answer

A valid statistical indicator needs a clearly stated concept and operational definition. Combining distinct ideas in one question may make interpretation difficult, while ambiguous response choices can introduce measurement error. Scientific and administrative evidence should remain separable.

Here the initial public assertion is: Dataville asks residents whether they are ‘satisfied and successful’ using one question and reports the percentage as measured wellbeing. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: A valid statistical indicator needs a clearly stated concept and operational definition. Combining distinct ideas in one question may make interpretation difficult, while ambiguous response choices can introduce measurement error. Scientific and administrative evidence should remain separable. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For question wording, definitions and measurement validity, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

In an imaginary survey 60 of 100 respondents select one broad answer. The 60% response cannot establish how many separately felt satisfied and successful.

Rainbolt reveals the changed evidence

Rainbolt splits the question into two and receives different distributions.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks whether all participants understand the wording under their circumstances.

Safe three-person tutorial and family activity

Rewrite fictional questions to avoid leading, double-barrelled and vague concepts.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville asks residents whether they are ‘satisfied and successful’ using one question and reports the percentage as measured wellbeing. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: A valid statistical indicator needs a clearly stated concept and operational definition. Combining distinct ideas in one question may make interpretation difficult, while ambiguous response choices can introduce measurement error. Scientific and administrative evidence should remain separable. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For question wording, definitions and measurement validity, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: In an imaginary survey 60 of 100 respondents select one broad answer. The 60% response cannot establish how many separately felt satisfied and successful. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt splits the question into two and receives different distributions. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks whether all participants understand the wording under their circumstances. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain question wording, definitions and measurement validity, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 05 — Frames, sampling weights and who the sample can represent

The Dataville example: A national report applies a street-interview sample to the whole country, although remote residents had no chance of selection.

The statistical question and immediate useful answer

Statistical inference depends on the sampling frame, selection probabilities, appropriate weights and assumptions. Some convenience samples cannot yield valid representative estimates merely by multiplying their response counts.

Here the initial public assertion is: A national report applies a street-interview sample to the whole country, although remote residents had no chance of selection. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: Statistical inference depends on the sampling frame, selection probabilities, appropriate weights and assumptions. Some convenience samples cannot yield valid representative estimates merely by multiplying their response counts. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For frames, sampling weights and who the sample can represent, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

A toy poll has 80 urban and 20 rural respondents while the fictional population is evenly split. The unweighted eighty-percent urban share reflects the sample, not the population composition.

Rainbolt reveals the changed evidence

Rainbolt changes the assumed population shares and the weighting requirement becomes visible.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks how people not reached can be represented fairly.

Safe three-person tutorial and family activity

Construct fictional sampling-frame cards and identify what needs qualified survey design.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: A national report applies a street-interview sample to the whole country, although remote residents had no chance of selection. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: Statistical inference depends on the sampling frame, selection probabilities, appropriate weights and assumptions. Some convenience samples cannot yield valid representative estimates merely by multiplying their response counts. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For frames, sampling weights and who the sample can represent, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: A toy poll has 80 urban and 20 rural respondents while the fictional population is evenly split. The unweighted eighty-percent urban share reflects the sample, not the population composition. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt changes the assumed population shares and the weighting requirement becomes visible. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks how people not reached can be represented fairly. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain frames, sampling weights and who the sample can represent, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 06 — Nonresponse, missing values and misleading zeros

The Dataville example: Dataville treats blank questionnaire fields as numeric zero and concludes that some families reported no income or household members.

The statistical question and immediate useful answer

A missing value and an observed zero are different. Collection quality, response burden, item nonresponse and documented imputation can affect statistics; a pupil should not invent a missing respondent’s private answer.

Here the initial public assertion is: Dataville treats blank questionnaire fields as numeric zero and concludes that some families reported no income or household members. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: A missing value and an observed zero are different. Collection quality, response burden, item nonresponse and documented imputation can affect statistics; a pupil should not invent a missing respondent’s private answer. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For nonresponse, missing values and misleading zeros, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

An imagined dataset contains 100 rows with 20 blanks. The 20% missing share is a data-quality measure, not proof that twenty people had zero of the quantity.

Rainbolt reveals the changed evidence

Rainbolt discovers blanks were concentrated among people using an inaccessible form.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks how source practices protect both fairness and confidentiality.

Safe three-person tutorial and family activity

Compare paper tables using blank, zero, unavailable and suppressed labels.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville treats blank questionnaire fields as numeric zero and concludes that some families reported no income or household members. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: A missing value and an observed zero are different. Collection quality, response burden, item nonresponse and documented imputation can affect statistics; a pupil should not invent a missing respondent’s private answer. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For nonresponse, missing values and misleading zeros, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: An imagined dataset contains 100 rows with 20 blanks. The 20% missing share is a data-quality measure, not proof that twenty people had zero of the quantity. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt discovers blanks were concentrated among people using an inaccessible form. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks how source practices protect both fairness and confidentiality. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain nonresponse, missing values and misleading zeros, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 07 — Statistical disclosure control and private records

The Dataville example: Dataville publishes a table identifying exact incomes and medical records for individual households under a claim of total transparency.

The statistical question and immediate useful answer

The Fundamental Principles protect individual data collected for statistical purposes. Public access to aggregate statistics and appropriate authorised research access differ from publishing personally identifiable microdata. Even small aggregates may be sensitive under certain conditions.

Here the initial public assertion is: Dataville publishes a table identifying exact incomes and medical records for individual households under a claim of total transparency. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: The Fundamental Principles protect individual data collected for statistical purposes. Public access to aggregate statistics and appropriate authorised research access differ from publishing personally identifiable microdata. Even small aggregates may be sensitive under certain conditions. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For statistical disclosure control and private records, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

A fictional district has four households and a displayed mean; if three values are public, the fourth could be inferred by subtraction. That is a toy disclosure risk, not real legal advice.

Rainbolt reveals the changed evidence

Rainbolt links the aggregate with another imaginary list and reidentifies a row.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks that participants retain dignity and a legitimate information boundary.

Safe three-person tutorial and family activity

Use nonsense identifiers and demonstrate inference risk without real personal data.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville publishes a table identifying exact incomes and medical records for individual households under a claim of total transparency. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: The Fundamental Principles protect individual data collected for statistical purposes. Public access to aggregate statistics and appropriate authorised research access differ from publishing personally identifiable microdata. Even small aggregates may be sensitive under certain conditions. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For statistical disclosure control and private records, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: A fictional district has four households and a displayed mean; if three values are public, the fourth could be inferred by subtraction. That is a toy disclosure risk, not real legal advice. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt links the aggregate with another imaginary list and reidentifies a row. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks that participants retain dignity and a legitimate information boundary. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain statistical disclosure control and private records, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 08 — Impartial methods and professional independence

The Dataville example: The fictional mayor asks analysts to change the reference year so the chart looks better and demands this be hidden.

The statistical question and immediate useful answer

Professional independence and scientifically grounded methods support trust. Changes to definitions or source years may be legitimate, but must be explained; politically convenient undisclosed manipulation is inconsistent with the UN principles.

Here the initial public assertion is: The fictional mayor asks analysts to change the reference year so the chart looks better and demands this be hidden. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: Professional independence and scientifically grounded methods support trust. Changes to definitions or source years may be legitimate, but must be explained; politically convenient undisclosed manipulation is inconsistent with the UN principles. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For impartial methods and professional independence, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

A toy series rises from 70 to 80 under one baseline; comparing 80 to a different prior-year value of 90 gives a different direction. The comparison must state its periods.

Rainbolt reveals the changed evidence

Rainbolt finds the published graph omitted a year that alters interpretation.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks how statisticians can communicate honestly while remaining accountable to the public.

Safe three-person tutorial and family activity

Write two transparent chart captions from the same fictional series.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: The fictional mayor asks analysts to change the reference year so the chart looks better and demands this be hidden. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: Professional independence and scientifically grounded methods support trust. Changes to definitions or source years may be legitimate, but must be explained; politically convenient undisclosed manipulation is inconsistent with the UN principles. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For impartial methods and professional independence, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: A toy series rises from 70 to 80 under one baseline; comparing 80 to a different prior-year value of 90 gives a different direction. The comparison must state its periods. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt finds the published graph omitted a year that alters interpretation. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks how statisticians can communicate honestly while remaining accountable to the public. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain impartial methods and professional independence, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 09 — Comparable classifications across countries and years

The Dataville example: Dataville’s education chart combines a country’s primary-school pupils with another’s total secondary enrolment and labels them directly comparable.

The statistical question and immediate useful answer

International classifications and harmonised concepts can enable meaningful comparison, but national rules, age systems, currency, units and reference periods vary. Similar titles do not guarantee matching denominators or content.

Here the initial public assertion is: Dataville’s education chart combines a country’s primary-school pupils with another’s total secondary enrolment and labels them directly comparable. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: International classifications and harmonised concepts can enable meaningful comparison, but national rules, age systems, currency, units and reference periods vary. Similar titles do not guarantee matching denominators or content. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For comparable classifications across countries and years, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

A toy table lists 100 primary and 100 secondary students. Equal totals do not establish identical education access or learning outcomes.

Rainbolt reveals the changed evidence

Rainbolt discovers one country counts part-time learners and the other does not.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks why public comparisons should not unfairly misrepresent communities.

Safe three-person tutorial and family activity

Build a fictional cross-country metadata comparison sheet.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville’s education chart combines a country’s primary-school pupils with another’s total secondary enrolment and labels them directly comparable. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: International classifications and harmonised concepts can enable meaningful comparison, but national rules, age systems, currency, units and reference periods vary. Similar titles do not guarantee matching denominators or content. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For comparable classifications across countries and years, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: A toy table lists 100 primary and 100 secondary students. Equal totals do not establish identical education access or learning outcomes. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt discovers one country counts part-time learners and the other does not. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks why public comparisons should not unfairly misrepresent communities. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain comparable classifications across countries and years, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 10 — Revisions, preliminary estimates and statistical truthfulness

The Dataville example: A first release estimates economic activity and later corrections arrive, but Dataville calls the revised figure proof statistics are inherently dishonest.

The statistical question and immediate useful answer

Statistical estimates can be revised when improved or fuller data become available. Preliminary, revised and final values should be labelled, explained and archived; a transparent correction may strengthen rather than destroy trust.

Here the initial public assertion is: A first release estimates economic activity and later corrections arrive, but Dataville calls the revised figure proof statistics are inherently dishonest. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: Statistical estimates can be revised when improved or fuller data become available. Preliminary, revised and final values should be labelled, explained and archived; a transparent correction may strengthen rather than destroy trust. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For revisions, preliminary estimates and statistical truthfulness, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

A toy index was released at 110 then revised to 108, a two-point downward change. The update alone does not determine the reason or quality of both vintages.

Rainbolt reveals the changed evidence

Rainbolt shows the earlier result had a different seasonal adjustment method.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks whether readers can find why the value changed and whether old citations should be corrected.

Safe three-person tutorial and family activity

Maintain a two-version paper table with sources, reasons and dates.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: A first release estimates economic activity and later corrections arrive, but Dataville calls the revised figure proof statistics are inherently dishonest. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: Statistical estimates can be revised when improved or fuller data become available. Preliminary, revised and final values should be labelled, explained and archived; a transparent correction may strengthen rather than destroy trust. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For revisions, preliminary estimates and statistical truthfulness, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: A toy index was released at 110 then revised to 108, a two-point downward change. The update alone does not determine the reason or quality of both vintages. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt shows the earlier result had a different seasonal adjustment method. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks whether readers can find why the value changed and whether old citations should be corrected. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain revisions, preliminary estimates and statistical truthfulness, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 11 — Mean, median, outliers and distributions

The Dataville example: Dataville reports average household wealth and says this amount describes the ordinary experience of every resident.

The statistical question and immediate useful answer

Means and medians summarise different features of a distribution. Extreme values and unequal population shares can make one headline average misleading; no single statistic captures every person’s circumstances.

Here the initial public assertion is: Dataville reports average household wealth and says this amount describes the ordinary experience of every resident. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: Means and medians summarise different features of a distribution. Extreme values and unequal population shares can make one headline average misleading; no single statistic captures every person’s circumstances. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For mean, median, outliers and distributions, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

For fictional values 1, 2, 3, 4 and 90, the mean is 20 and the median is 3. Both are correct for different questions.

Rainbolt reveals the changed evidence

Rainbolt adds another very high value and the mean changes more than the median.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks how public data can acknowledge differences without disclosing private lives.

Safe three-person tutorial and family activity

Calculate fictional summary statistics and explain what each one leaves out.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville reports average household wealth and says this amount describes the ordinary experience of every resident. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: Means and medians summarise different features of a distribution. Extreme values and unequal population shares can make one headline average misleading; no single statistic captures every person’s circumstances. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For mean, median, outliers and distributions, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: For fictional values 1, 2, 3, 4 and 90, the mean is 20 and the median is 3. Both are correct for different questions. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt adds another very high value and the mean changes more than the median. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks how public data can acknowledge differences without disclosing private lives. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain mean, median, outliers and distributions, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 12 — Population versus geographic denominators

The Dataville example: Dataville colours 40% of a map and claims 40% of residents receive the service, without knowing where people live.

The statistical question and immediate useful answer

Statistics linked to geography require appropriate reference areas and a clear relationship between area, households and individuals. The UN Global Statistical Geospatial Framework supports integration but does not erase privacy or denominator questions.

Here the initial public assertion is: Dataville colours 40% of a map and claims 40% of residents receive the service, without knowing where people live. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: Statistics linked to geography require appropriate reference areas and a clear relationship between area, households and individuals. The UN Global Statistical Geospatial Framework supports integration but does not erase privacy or denominator questions. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For population versus geographic denominators, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

A fictional map has 100 cells, 40 serviced, but 80% of imaginary residents occupy the unserviced cells. The 40% area share cannot equal population coverage.

Rainbolt reveals the changed evidence

Rainbolt reveals the most densely populated region was excluded.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks how accessibility and spatial inequality can be represented accurately.

Safe three-person tutorial and family activity

Overlay invented area and population cards without using real household addresses.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville colours 40% of a map and claims 40% of residents receive the service, without knowing where people live. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: Statistics linked to geography require appropriate reference areas and a clear relationship between area, households and individuals. The UN Global Statistical Geospatial Framework supports integration but does not erase privacy or denominator questions. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For population versus geographic denominators, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: A fictional map has 100 cells, 40 serviced, but 80% of imaginary residents occupy the unserviced cells. The 40% area share cannot equal population coverage. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt reveals the most densely populated region was excluded. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks how accessibility and spatial inequality can be represented accurately. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain population versus geographic denominators, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 13 — Quality assurance across the complete statistical workflow

The Dataville example: Dataville tests a dataset only when it is published, leaving collection errors, inconsistent transformations and undocumented assumptions unexamined.

The statistical question and immediate useful answer

The UN NQAF provides a framework for quality across design, collection, processing, analysis and dissemination. Statistical quality involves relevance, accuracy, timeliness, comparability, coherence, access and systematic governance, not one checklist tick.

Here the initial public assertion is: Dataville tests a dataset only when it is published, leaving collection errors, inconsistent transformations and undocumented assumptions unexamined. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: The UN NQAF provides a framework for quality across design, collection, processing, analysis and dissemination. Statistical quality involves relevance, accuracy, timeliness, comparability, coherence, access and systematic governance, not one checklist tick. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For quality assurance across the complete statistical workflow, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

In a fictional quality record twenty fields are expected and fifteen documented. Seventy-five-percent paperwork completeness cannot certify statistical fitness for policy.

Rainbolt reveals the changed evidence

Rainbolt discovers an automated merge duplicated a group of survey entries.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks how public services can admit limitations before flawed indicators drive decisions.

Safe three-person tutorial and family activity

Draw a paper source-to-publication workflow and mark quality gates.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville tests a dataset only when it is published, leaving collection errors, inconsistent transformations and undocumented assumptions unexamined. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: The UN NQAF provides a framework for quality across design, collection, processing, analysis and dissemination. Statistical quality involves relevance, accuracy, timeliness, comparability, coherence, access and systematic governance, not one checklist tick. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For quality assurance across the complete statistical workflow, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: In a fictional quality record twenty fields are expected and fifteen documented. Seventy-five-percent paperwork completeness cannot certify statistical fitness for policy. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt discovers an automated merge duplicated a group of survey entries. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks how public services can admit limitations before flawed indicators drive decisions. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain quality assurance across the complete statistical workflow, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 14 — SDG indicators, targets and missing data

The Dataville example: A school declares a UN Sustainable Development Goal achieved because one national indicator has a rising line for two years.

The statistical question and immediate useful answer

SDG goals, targets and indicators are related but distinct. Each indicator has metadata, custodians, populations and timeframes; one improving series does not establish an entire target or goal. Data gaps should remain visible.

Here the initial public assertion is: A school declares a UN Sustainable Development Goal achieved because one national indicator has a rising line for two years. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: SDG goals, targets and indicators are related but distinct. Each indicator has metadata, custodians, populations and timeframes; one improving series does not establish an entire target or goal. Data gaps should remain visible. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For sdg indicators, targets and missing data, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

A toy goal uses four indicators and one is improving. One-quarter improving in that artificial group cannot be called 25% completion of the goal.

Rainbolt reveals the changed evidence

Rainbolt finds another indicator lacks recent observations.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks how to speak honestly about progress and inequality without claiming universal outcomes.

Safe three-person tutorial and family activity

Classify imaginary goal, target and indicator cards by hierarchy and metadata.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: A school declares a UN Sustainable Development Goal achieved because one national indicator has a rising line for two years. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: SDG goals, targets and indicators are related but distinct. Each indicator has metadata, custodians, populations and timeframes; one improving series does not establish an entire target or goal. Data gaps should remain visible. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For sdg indicators, targets and missing data, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: A toy goal uses four indicators and one is improving. One-quarter improving in that artificial group cannot be called 25% completion of the goal. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt finds another indicator lacks recent observations. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks how to speak honestly about progress and inequality without claiming universal outcomes. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain sdg indicators, targets and missing data, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 15 — AI, automated data pipelines and independent checking

The Dataville example: Dataville uses an AI system to classify thousands of records and assumes no human source review is needed because the charts look consistent.

The statistical question and immediate useful answer

Automation may improve speed and consistency while introducing bias, coding errors, opaque transformations and privacy risks. The UN handbook’s 2025/A update addresses data ecosystems and governance. Responsible statistical systems document methods and retain professional review.

Here the initial public assertion is: Dataville uses an AI system to classify thousands of records and assumes no human source review is needed because the charts look consistent. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: Automation may improve speed and consistency while introducing bias, coding errors, opaque transformations and privacy risks. The UN handbook’s 2025/A update addresses data ecosystems and governance. Responsible statistical systems document methods and retain professional review. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For ai, automated data pipelines and independent checking, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

A toy automated table has 1,000 entries, with 40 confirmed classification errors in a checked subset of 100. The 40% subset-error rate cannot be assumed to apply to all 1,000 without proper design.

Rainbolt reveals the changed evidence

Rainbolt reveals the review sample contained only the most difficult cases.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks about transparency, individual privacy and communities harmed by wrong classifications.

Safe three-person tutorial and family activity

Use invented data-processing cards and a human-review role without automated personal profiling.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville uses an AI system to classify thousands of records and assumes no human source review is needed because the charts look consistent. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: Automation may improve speed and consistency while introducing bias, coding errors, opaque transformations and privacy risks. The UN handbook’s 2025/A update addresses data ecosystems and governance. Responsible statistical systems document methods and retain professional review. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For ai, automated data pipelines and independent checking, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: A toy automated table has 1,000 entries, with 40 confirmed classification errors in a checked subset of 100. The 40% subset-error rate cannot be assumed to apply to all 1,000 without proper design. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt reveals the review sample contained only the most difficult cases. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks about transparency, individual privacy and communities harmed by wrong classifications. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain ai, automated data pipelines and independent checking, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Mission 16 — Statistical archives and trust beyond the founding team

The Dataville example: Dataville’s original chief statistician retires, leaving colourful dashboards but no definitions, privacy documentation or record of revisions.

The statistical question and immediate useful answer

Durable statistical institutions require trustworthy metadata, professional staff, legal independence, reproducible methods, training, data protection and accessible correction processes. A chart without context becomes less useful as generations and technologies change.

Here the initial public assertion is: Dataville’s original chief statistician retires, leaving colourful dashboards but no definitions, privacy documentation or record of revisions. The question becomes clearer when the reader identifies the intended target population and the specific public service or research goal. The central method is: Durable statistical institutions require trustworthy metadata, professional staff, legal independence, reproducible methods, training, data protection and accessible correction processes. A chart without context becomes less useful as generations and technologies change. That is more useful than merely declaring the chart impressive.

Where the source came from and how it can fail

For statistical archives and trust beyond the founding team, trace collection, definitions, the responsible statistical producer and the eventual public claim. Some parts of that chain can be verified through the original institutional source, while Dataville’s population and figures remain invented. The group should mark an uncertain data step instead of creating an imagined professional approval.

Professional independence and public clarity

Work through the mathematics, then examine the denominator

Twelve of sixteen imaginary datasets have definition, source, privacy classification and review date. Seventy-five percent documentation does not prove results are accurate.

Rainbolt reveals the changed evidence

Rainbolt finds a popular graphic that uses a retired definition without a warning.

Wintour’s confidentiality, inclusion and user perspective

Wintour asks whether future learners can find the right version and meaning of each statistic.

Safe three-person tutorial and family activity

Give a new pupil a fictional data dictionary and ask which claims remain unverified.

Four worked source and reasoning cards

Identify what the reported statistic truly counts

The headline says: Dataville’s original chief statistician retires, leaving colourful dashboards but no definitions, privacy documentation or record of revisions. Specify a measurable concept, correct source, intended population, period and one inference the reader should not make from the supplied facts.

A strong answer returns to the specific issue: Durable statistical institutions require trustworthy metadata, professional staff, legal independence, reproducible methods, training, data protection and accessible correction processes. A chart without context becomes less useful as generations and technologies change. It distinguishes invented Dataville figures from principles or statistical data published by competent institutions. The school should reward careful limitations rather than a dramatic claim about a country, demographic group or individual.

For statistical archives and trust beyond the founding team, rotate the source historian, maths auditor and public-user reviewer. Each learner paraphrases another’s strongest question before the group publishes a short corrected explanation. No real survey responses, family income, health information or private business records are required.

An exact calculation may answer the wrong question

The fictional arithmetic is: Twelve of sixteen imaginary datasets have definition, source, privacy classification and review date. Seventy-five percent documentation does not prove results are accurate. Recalculate when an input changes, and explain what the numerator and denominator actually represent rather than claiming a public outcome.

A new version creates a necessary correction

Rainbolt reports: Rainbolt finds a popular graphic that uses a retired definition without a warning. State which chart caption must change and how the statistical office would make the revision source and date visible.

The respondent or reader who was overlooked

Wintour asks: Wintour asks whether future learners can find the right version and meaning of each statistic. Give a fair, privacy-respecting concern, one needed source check and an appropriate method for public communication.

Successor record: A new pupil can explain statistical archives and trust beyond the founding team, reconstruct the invented measure, identify the official statistical source and see why the original public interpretation needed correction.

Eight-week official-statistics literacy course

Pair two missions each week, starting with what makes a statistic official, moving through data collection, sampling, confidentiality and methodological integrity, and ending with public evidence, AI and archival continuity. All figures are fictional and activities are restricted to source reading, paper tables and critical discussion.

Week 1: Official statistics are public evidence, not government marketing and National statistical office and specialised public data units

The first session examines official statistics are public evidence, not government marketing, then compares it with national statistical office and specialised public data units. Have pupils formulate the actual public question before choosing whether a census, survey, register or model may be relevant. The second topic should reveal a data or rights issue the original chart omitted.

Allocate ten minutes to an official UN or Singapore statistical reference, ten to a wholly imaginary source table, ten to a labelled calculation and ten to correcting a chart caption after a changed assumption. Primary pupils may classify types of records; Secondary students should compute means, medians or appropriately bounded response percentages.

Three students rotate statistical producer, arithmetic auditor and ordinary reader. Ask each to state the strongest limitation of the first interpretation and give one source for the remaining question. The tutor should never solicit private information about children or their families.

Week 2: Census, sample survey and administrative register and Question wording, definitions and measurement validity

The first session examines census, sample survey and administrative register, then compares it with question wording, definitions and measurement validity. Have pupils formulate the actual public question before choosing whether a census, survey, register or model may be relevant. The second topic should reveal a data or rights issue the original chart omitted.

Week 3: Frames, sampling weights and who the sample can represent and Nonresponse, missing values and misleading zeros

The first session examines frames, sampling weights and who the sample can represent, then compares it with nonresponse, missing values and misleading zeros. Have pupils formulate the actual public question before choosing whether a census, survey, register or model may be relevant. The second topic should reveal a data or rights issue the original chart omitted.

Week 4: Statistical disclosure control and private records and Impartial methods and professional independence

The first session examines statistical disclosure control and private records, then compares it with impartial methods and professional independence. Have pupils formulate the actual public question before choosing whether a census, survey, register or model may be relevant. The second topic should reveal a data or rights issue the original chart omitted.

Week 5: Comparable classifications across countries and years and Revisions, preliminary estimates and statistical truthfulness

The first session examines comparable classifications across countries and years, then compares it with revisions, preliminary estimates and statistical truthfulness. Have pupils formulate the actual public question before choosing whether a census, survey, register or model may be relevant. The second topic should reveal a data or rights issue the original chart omitted.

Week 6: Mean, median, outliers and distributions and Population versus geographic denominators

The first session examines mean, median, outliers and distributions, then compares it with population versus geographic denominators. Have pupils formulate the actual public question before choosing whether a census, survey, register or model may be relevant. The second topic should reveal a data or rights issue the original chart omitted.

Week 7: Quality assurance across the complete statistical workflow and SDG indicators, targets and missing data

The first session examines quality assurance across the complete statistical workflow, then compares it with sdg indicators, targets and missing data. Have pupils formulate the actual public question before choosing whether a census, survey, register or model may be relevant. The second topic should reveal a data or rights issue the original chart omitted.

Week 8: AI, automated data pipelines and independent checking and Statistical archives and trust beyond the founding team

The first session examines ai, automated data pipelines and independent checking, then compares it with statistical archives and trust beyond the founding team. Have pupils formulate the actual public question before choosing whether a census, survey, register or model may be relevant. The second topic should reveal a data or rights issue the original chart omitted.

Twenty practical questions about public statistics

What makes a statistic official? It is produced under an authorised statistical system with defined professional principles and responsibilities.
Are official statistics merely government advertising? No. The UN principles emphasise impartiality and professional independence.
When were the UN Fundamental Principles endorsed by the General Assembly? In 2014.
Is a census the same as a voluntary sample? No, coverage and inference differ.
Can a large dataset still have bias? Yes, if it systematically excludes relevant people or contains measurement errors.
Does response percentage establish representativeness? No, sampling and nonresponse matter.
Why should missing be different from zero? A blank indicates lack of a recorded value, whereas zero is an observed quantity.
Can statistics be revised honestly? Yes, improved data and transparent methods can require corrections.
Does a high average describe every household? No. Means, medians and distributions differ.
Can administrative records have other primary purposes? Yes, they are often designed for services rather than statistical measurement.
What is statistical confidentiality? Protecting individually collected information from inappropriate disclosure and use.
Can aggregate statistics reveal identities? In some small or linked datasets, indirect disclosure risks remain.
Why compare indicators with matching denominators? Otherwise the figures may measure different groups or outcomes.
What is UN NQAF? A professional framework for quality assurance in official statistics adopted in 2019.
How is Singapore’s statistical system organised? The DOS and Research and Statistics Units operate under a decentralised framework with national coordination.
What changed in UNSD’s 2025/A handbook? It includes an updated chapter on national data ecosystems and governance.
Does a graph’s unitless axis promote sound analysis? No, units and reference periods must be clear.
Can one SDG indicator prove the whole target achieved? No, targets and indicators must be distinguished.
Can AI substitute for statistical review? No, professional governance and transparent methods remain essential.
What should future statistical organisations preserve? Source methods, quality, confidentiality, revisions and usable public access.

Official-statistics glossary

Official statistics — public data produced by competent statistical authorities under a defined framework
Statistical system — network of producers, coordinators, standards and sources of official statistics
National statistical office — institution responsible for coordinating or producing selected official statistics
Census — enumeration of a population under a specified reference frame
Sample survey — collection of information from a subset under a sampling design
Administrative data — records originally collected for programme or service purposes
Sampling frame — list or mechanism defining which population elements may be selected
Nonresponse — missing participation or answers under a survey method
Sampling weight — factor used under appropriate statistical inference
Statistical confidentiality — restrictions protecting individual data and its proper use
Quality assurance — organised methods for preserving statistical fitness for purpose
Metadata — information describing the definition, sources, methods, period and limits of data
Statistical revision — documented change after improved data or methods
Mean — arithmetic average under a defined dataset
Median — central ordered value under a dataset
Denominator — total or reference population used in a ratio
Statistical bias — systematic distortion relative to an intended measurement
Disclosure control — methods and safeguards reducing risks of revealing identifiable data
Indicator — defined statistical measure used to monitor a phenomenon
Provenance — traceable origin, date, calculation and institutional responsibility of a statistic

Impartiality

Professional dissemination of statistics without favouring a preferred policy conclusion.

Ask a pupil to use ‘Impartiality’ in a precise fictional example, then rewrite a sentence that overstates what a statistic can establish. A source and a denominator should accompany numerical claims.

Timeliness

How quickly information becomes available relative to the phenomenon and user need.

Ask a pupil to use ‘Timeliness’ in a precise fictional example, then rewrite a sentence that overstates what a statistic can establish. A source and a denominator should accompany numerical claims.

Comparability

Suitability of statistics for meaningful comparisons under aligned definitions.

Ask a pupil to use ‘Comparability’ in a precise fictional example, then rewrite a sentence that overstates what a statistic can establish. A source and a denominator should accompany numerical claims.

Representativeness

Extent to which data permit valid inference about an intended population under stated methods.

Ask a pupil to use ‘Representativeness’ in a precise fictional example, then rewrite a sentence that overstates what a statistic can establish. A source and a denominator should accompany numerical claims.

Data version

Documented edition of a dataset or release.

Ask a pupil to use ‘Data version’ in a precise fictional example, then rewrite a sentence that overstates what a statistic can establish. A source and a denominator should accompany numerical claims.

Statistical literacy

Ability to understand, assess and communicate numerical claims in context.

Ask a pupil to use ‘Statistical literacy’ in a precise fictional example, then rewrite a sentence that overstates what a statistic can establish. A source and a denominator should accompany numerical claims.

Featured media and the concluding public evidence test

The featured image is existing eduKateSG educational artwork of researchers checking an abstract dataset and validating records in a university data workflow. It is conceptual rather than a photograph of Singapore’s Department of Statistics or a real national statistical release. It illustrates verification work without implying access to private identifiable records.

All Dataville people, survey values, records and results are imaginary. The UN Fundamental Principles, UN NQAF, Singapore DOS guidance and the UNSD Handbook are real institutional materials whose authority, dates and scope must be preserved. No example is an actual audited government statistic or authorisation to release individual data.

Final Dataville question: When founding statisticians and software developers leave, can their successors still explain what was counted, protect individual information, correct a flawed graph and make reliable evidence available to the public? That is the enduring value of an official-statistics civilisation.