Back to Top

 Skip navigation

What is a Classification?

A classification is a way to group information into categories and sets of related categories in a meaningful, systematic, and logical format to make large volumes of data more manageable, presentable, and understandable. 

The OECD Glossary of Statistical Terms states that “Standard classifications are those that follow prescribed rules and are generally recommended and accepted. They aim to ensure that information is classified consistently regardless of the collection, source, point of time etc.”

Statistical classifications are a key requirement for the production of reliable, comparable, and methodologically sound statistics. Classifications are used to organise information into categories with other similar pieces of information. They are therefore an important part of any data standard.  

Features of a classification

Classifications are used in most parts of the statistical cycle and need to be carefully considered when designing any data or statistical system. A classification organises or groups data according to its similarities. It consists of codes and descriptors and allows survey responses to be put into meaningful categories in order to produce useful data. A classification is a critical tool for anyone developing statistical surveys. It is a framework which both simplifies the topic being studied and makes it easy to categorise all data or responses received.

Ideally, standard classifications contain categories that are mutually exclusive and exhaustive, i.e. any given response or data point can be matched to only one category at a particular level of the classification structure and that every possible response is anticipated and catered for.

It is generally accepted that a standard statistical classification will usually meet a number of requirements which are outlined below.

Key features of a statistical classification

  • Exhaustive categories – All potential responses need to fit into the classification structure somewhere, and classifications need to be updated to cater for changes over time
  • Unambiguous, precise, and meaningful descriptors for categories – The content of each category in the classification should be clearly defined
  • Conceptually sound – The classification should have a clear conceptual basis and a logical structure
  • Operationally feasible – It should be achievable to collect data using the categorisations - there is no point in having a classification that cannot be implemented in practice
  • Statistically robust – The classification should be able to be used for a reasonable timeframe without major revisions
  • Internationally comparable – The classification should be comparable with related international standard classifications
  • Categories should be mutually exclusive at each level of the classification - Each unit or response should only be classified to one category, and it should be possible to classify all units to a category in the classification (avoiding overlaps or double-counting)

CSO Classification Process

In Ireland, international statistical classifications are widely utilised across many key statistical outputs to ensure data are comparable across European and world statistics. These standard classifications are used for collecting, processing, and outputting data on a particular topic. These international standards are developed following extensive research of the topic in close consultation with stakeholders.

While many international standard classifications exist, it is often necessary for individual National Statistical Institutes, or indeed specific statistical outputs, to adapt or modify these in order to cater for the needs of their own country or to avoid statistical disclosure. There is also a need to develop further classifications for national or survey specific purposes where there is no wider European reporting requirement. Within the Central Statistics Office, many such classifications exist, with ongoing work being done to apply rules and procedures to align these with the principles of standard classifications. When developing a new data collection or survey, it is always advisable to seek guidance from the CSO Classifications Management Unit on the availability or adaptability of existing classifications or the formulation of new classifications.