|Page tools: Print Page Print All RSS Search this Product|
Census Classifications and Supplementary Codes
Most classifications in the Census Dictionary contain ‘Supplementary codes’. Supplementary codes are used to code responses that otherwise do not fit into the main classification.
This category is used when no response has been provided to a particular question on the Census form. This differs to surveys that usually do not have a "not stated" category because they determine answers though interviewer assistance or may impute results on the few occasions where they do not receive a response.
The Census imputes information for the variables for age (AGEP), sex (SEXP), registered marital status (MSTP) and usual residence (PURP). All other Census variables have not stated categories. The ampersand symbol ‘&’ is used to denote ‘Not stated’ in 2016.
An exception to this is for family and household relationships. If all the people on the form did not state their relationships, it is coded as a ‘Not classifiable’ household. If only some people in the household did not state relationships, then their relationship will be inferred from other information on the form, or else coded to ‘otherwise unrelated’.
Treatment of ‘Not Stated’ Responses
When calculating percentages with Census data, you may choose to include or exclude the ‘Not stated’ responses depending on the Census variables you’re examining. When you exclude ‘Not stated’ responses prior to the calculation of percentages, you are effectively distributing the ‘Not stated’ results across the remaining categories. In doing so, you are assuming the stated responses are representative of the ‘Not stated’ responses.
This may not necessarily be a safe assumption to make: particular population groups may be more likely not to respond to certain questions than other population groups. For example, low-income earners may be less likely to respond to the personal income question than high-income earners. If you exclude the ‘Not stated’ category in such instances, make sure you explain that their exclusion may have had an impact on the information presented.
What is most important is that you clearly specify when ‘Not stated’ responses have been included or excluded from the table or analysis.
The ‘Not applicable’ category exists for questions which do not apply to all persons and so, for these people, no response is required. For example, the question asking for a person’s Year of Arrival in Australia is not applicable to people born in Australia – in this case, all Australian-born people will be classified as ‘Not applicable’. Similarly, the question asking a person’s occupation is ‘Not applicable’ for persons aged under 15 years, as is ANC2P for people who only provided one response to the ancestry question.
In the Census Dictionary the classification information states explicitly which population groups have been classified as ‘Not applicable’. The ‘@’ symbol is used for ‘Not applicable’ in 2016.
People who usually live in another country and were visiting Australia for less than one year are recorded in Census data as an ‘Overseas visitor’. Only information on age, sex and registered marital status is recorded for overseas visitors; all other information is coded to the ‘Overseas visitors’ supplementary category for each variable. The letter ‘V’ is used to denote overseas visitors. Many users wish to exclude overseas visitors from tables they have built. This can be done by using any variable (except AGEP, SEXP and MSTP) by grouping all responses together and excluding the overseas visitor category.
Indigenous Status is commonly used to exclude overseas visitors.
This category captures responses not clear enough to be categorised elsewhere. For example, if someone gave their birthplace as Earth or Soviet Union, this would be coded to ‘Inadequately described’.
Not Elsewhere Classified (nec)
This is used as a "dump code" for legitimate responses that do not have an individual listing in the classification. For example, if someone gave their ancestry as Greenlandic, it is coded to ‘2499 Northern European, nec’ along with Faeroese and Saami.
Not Further Defined (nfd)
This code is used when enough information exists to partially code information, but there is not enough information to code it to the most detailed category in the classification. For example, if someone listed their birthplace as ‘Yugoslavia’ it is coded to ‘South Eastern Europe, nfd’ because there is not enough information to know whether the person should be coded to Serbia, Montenegro, etc.
When calculating percentages in classifications which use the ‘Inadequately described’, ‘nec’, or ‘nfd’ supplementary codes, you should treat these data categories as standard categories – that is, don’t exclude and distribute them the way you might with ‘Not stated’ categories.
These documents will be presented in a new window.