Standards and classifications
A statistical standard is a set of rules used to standardise the way data are collected and statistics are produced. They provide information about data collected on a particular topic that assists in the understanding and interpretation of that data.
Statistical standards are the approved versions of how to:
- define the underlying concept
- define the specific variables of interest
- collect data
- coding structure
- statistical units
- recommended question modules
- process the data
- standard editing
- present the data
- output categories
- interpret the data
Classifications are used to collect and organise information into categories with other similar pieces of information. They are an important part of any standard.
Classifications are used in most parts of the statistical cycle including:
- data collection
- data processing
- data presentation
- data analysis
Classifications should be exhaustive, and mutually exclusive.
For example, Australia, France, Japan, and Vanuatu are some of the categories from the ‘Standard Australian Classification of Countries’ (SACC). This classification is used in the ‘Country of Birth Standard’ where each response for a person's country of birth will belong in one and only one category within the classification.
Importance of standards in statistics
Standards are used to ensure that data about the same characteristic are collected and communicated in the same way every time. This enables data from different sources to be compared on a consistent basis and enables meaningful comparisons to be made over time.
There are four key advantages from having widespread use of approved statistical standards:
- ensure the quality of statistical outputs
- creates a meaningful statistical picture of society and economy
- reduces costs
- improves transparency
Uses of statistical standards
Standards are used:
- internally in the Australian Bureau of Statistics to ensure data are produced to a consistent quality and in a consistent manner over time and across collections
- nationally by those producing statistics to assist with the integration of data from various sources
- internationally to comply with international reporting obligations and encourage data comparability between countries.
Comparability is the ability to validly compare statistics that have been collected over time, or from different sources.
For example, in the standard definitions used for ABS labour force collections, a person must be actively seeking employment and available to start work to be classified as unemployed. If the definition of unemployed is changed to include people who were not actively seeking work and/or not available to start work, the two datasets would not be comparable as the data has not been collected using common definitions. It would be difficult for users to determine whether a change in the reported unemployment rate was due to a change in definitions used or a reflection of what is happening in the labour market.