← Back to catalogue
Published

statistics

vr.tr.statistics · thing-q12483

Let an agent handle statistics by method, data source, interpretation and communication, citing official statistics and avoiding misuse.

Thing Registry Information and virtual systems INF.KNW
Bundle → Layer → Finding → QuestionsClassifiersWhat it isWhy it existsDistinguishing featuresWhat robots and AI may and may not doMoral aspectsOwnersLinks to other meta-modelsIdentity and identifiersDirect propertiesRecognitionCapabilities and actionsHazards and failure modesStandards and interfacesContext of useSources

Bundle → Layer → Finding → Questions Filled

4 bundles · 8 layers · 8 findings · 16 questions

Data Sources.

Source

Official data.

Source

Data source.

  1. Which official or peer-reviewed source provides this statistic? provenance
  2. For which period and population? definition

Quality

Methods.

Quality

Data quality.

  1. How was the data collected and how large is the sample? measurement
  2. What are the known limitations? boundary
Analysis Methods.

Method

Choice.

Method

Method choice.

  1. Which statistical method suits this question and data? action
  2. What assumptions does it make? boundary

Uncertainty

Intervals.

Uncertainty

Uncertainty.

  1. What is the margin of error or confidence interval? measurement
  2. Is the result statistically and practically significant? boundary
Interpretation Meaning.

Causation

Correlation.

Causation

Correlation and causation.

  1. Does the evidence support causation or only correlation? boundary
  2. What design would test causation? definition

Misleading

Red flags.

Misleading

Misleading statistics.

  1. Is the chart or claim misleading, for example by truncated axes? boundary
  2. How can it be presented fairly? action
Learning Skills.

Literacy

Understanding.

Literacy

Statistical literacy.

  1. How can this statistic be explained in plain language? action
  2. Would natural frequencies help? action

Tools

Software.

Tools

Tools.

  1. Which software can perform this analysis? provenance
  2. Is it reproducible? boundary

Classifiers Filled

Family
Thing Registry
Category
Information and virtual systems
Entry kind
thing
Plane
INF
Domain
INF.KNW
Other names and narrower kinds
census data, statistical data, funnel analysis, statistical hydrology, forensic statistics, branch of statistics, circular analysis, demographic statistics, geostatistics, health statistics, statistical machine translation, museum statistics

What it is Filled

The discipline concerned with collecting, analysing, interpreting and presenting data, including descriptive and inferential statistics, probability models and official statistics produced by national agencies.

Why it exists Filled

Let an agent handle statistics by method, data source, interpretation and communication, citing official statistics and avoiding misuse.

Distinguishing features Filled

  • Studies data and uncertainty
  • Descriptive and inferential
  • Official statistics are regulated
  • Easily misused

What robots and AI may and may not do Filled

Must not

  • Report results selectively or tune analyses until they look significant.
  • Present correlation as causation without evidence.
  • Re-identify individuals from anonymised statistical data.
  • Fabricate or alter data.

Only with a human decision

  • Publishing official or research statistics.
  • Using statistical models to decide about individuals.

May

  • Analyse data the user is entitled to use, stating methods, assumptions and uncertainty.
  • Check statistical claims against their sources.

Moral aspects Filled

  • Statistics shape policy; errors and manipulation mislead the public.
  • Data about groups can stigmatise them or expose individuals.

Who is affected

  • People in the data
  • The public
  • Decision makers

Owners Filled

Steward

Nobody owns the discipline; statistical offices answer for official statistics.

Master systems

  • Official statistics databases

Links to other meta-models Filled

parent

  • Q395 - registry parent class
  • Q816264 - registry parent class

related

  • formal science - category
  • probability - foundation
  • national statistics office - official source
  • cognitive bias - misinterpretation

What else AI and robots need to interact with it Filled

Identity and identifiers required Filled

  • Vercy registry: vr.tr.statistics
  • Wikidata: Q12483 (https://www.wikidata.org/wiki/Q12483)

Direct properties not applicable Not applicable

Not applicable

Plane INF: no invented physical properties.

Recognition optional Filled

  • Means, medians, confidence intervals
  • Official statistics releases
  • A single statistic is one number, not the discipline
  • Tables, charts, models and reports.

Capabilities and actions required Filled

  • find official statistics
  • choose appropriate methods
  • interpret uncertainty and significance
  • spot misleading statistics

Hazards and failure modes required Filled

  • Misleading charts
  • P-hacking and misuse of p-values
  • Confusing correlation and causation

Standards and interfaces required Filled

  • UN Fundamental Principles of Official Statistics
  • European Statistics Code of Practice
  • ISO 3534 vocabulary

Context of use required Filled

  • Evidence-based decisions in science, government and business.
  • descriptive statistics
  • inferential statistics
  • Bayesian statistics
  • official statistics
  • applied fields such as forensic statistics

Sources Filled

  1. Wikidata item Q12483: statistics - identity and sense of the item
  2. Wikipedia: Statistics - general description of the item

Open questions

  • Should statistical methods be separate entries?
  • How should official statistics be linked?
  • How should uncertainty be communicated?

Machine files

Card as JSON Specification Card backlog vercy.card/1

Provenance

thing registry research (pass 2) · unreviewed

Built from: models/things/publications/thing-q12483/spec.json