## Statistical and Data Standards

## Are standards compulsory?

Each standard must include a statement regarding the degree to which its application is compulsory: departmental, recommended or program specific. More details can be found at Is your standard compulsory?

## How do we define standards?

Standards are comprised of statistical standards which in turn are defined by statistical data standards and statistical metadata standards. More details can be found at How do we define standards?

## What is a survey?

A survey is an activity in which information is collected for some units of the target population.

## What is a questionnaire?

A questionnaire is a research instrument that consists of a set of questions for the purpose of gathering information from respondents through survey or statistical study.

## What is a variable?

A variable is a characteristic of a statistical unit being observed that may assume more than one of a set of values to which a numerical measure or a category from a classification can be assigned.

## What is a statistical unit?

A statistical unit refers to the unit of observation or measurement for which data are collected or derived. It is aligned with what is generally referred to as the unit type, namely a class or group of units based on a single characteristic.^{Footnote 1}

## What is a statistical classification?

A set of categories which may be assigned to one or more variables registered in a statistical program and used in the production and dissemination of statistics. The categories at each level of the classification structure must reflect a well-defined universe, cover all possible elements in the universe (exhaustiveness), have mutually exclusive classes, be rectangular and hierarchical (in the case of classifications), be comparable to other classifications, have categories that are empirically significant, be organized around one or few principles, contains groupings that are meaningful to users and be widely adopted.^{Footnote 2}

## Classification variants

## What is a classification variant?

Variants are commonly of three kinds. These have been named **extension variants**, **aggregate variants** or **regrouping variants**. There could exist other types of variants. A particular variant could include elements from more than one of these variant types.^{Footnote 3}

## What is an extension variant?

An extension variant is a Statistical Classification that extends the base Statistical Classification with one or several new Levels at the bottom, creating a new lowest Level. An extension variant thus adds new lower Levels to the base Statistical Classification but does not otherwise alter its original.^{Footnote 3}

## What is an aggregate variant?

An aggregate variant is a Statistical Classification that groups the categories of a linear Statistical Classification to create one or several aggregate level(s), thus creating a hierarchy.^{Footnote 3}

## What is a regrouping variant?

A regrouping variant is a Statistical Classification that introduces additional or alternative aggregate levels by regrouping categories of the base statistical classification. Two types of regrouping variants have been identified:

- Regrouping variants which do not violate the structure of the base Statistical Classification: This type of regrouping variant introduces a new level or new levels on top of, or in between existing Levels of a hierarchical Statistical Classification without otherwise altering the original structure of the hierarchy. This regrouping variant consists of all classification Levels of the base Statistical Classification plus the new variant Level(s). The parent Level (if any) of the new variant Level can be either another variant Level or a Level from the base Statistical Classification.
- Regrouping variants which violate the structure of the base Statistical Classification: This type of regrouping variant introduces a new Level or new Levels on top of any but the topmost Level of a hierarchical Statistical Classification by regrouping categories of the base Statistical Classification in a way which violates its original order and structure. This regrouping variant consists of all classification Levels of the base Statistical Classification below the new variant Level(s) plus the new variant Level(s). In such a regrouping variant, a new variant Level cannot have a base Statistical Classification Level as parent Level.
^{Footnote 3}