A student is doing historical research with census data, and she found a few DAs for which we can't find any data through the archived census profile on the EFT or from other products that share census data such as the Census Analyser. Sometimes with historical research the students fail to realize that the boundaries have changed, which requires the help of the correspondence files, but these DAs definitely exist in 2001 as shown in boundary files or reference maps.
For example, DA 12090310 is located right in Halifax. But we can't find any data on it! In the 2001 Beyond 20/20 profiles, we jump straight from 12090309 to 12090311. So where can we find some sort of explanation about why data wasn't released for that DA and other DAs, preferably with lists of suppressed DAs for 2001 and EAs for older years?
Question 2In some cases, only some subsets of data are available for select DAs. For example, in 2006, 12090292 has the total number of census families and many other (but not all) variables, but age counts (?!) seem to be unavailable. The student would love some sort of explanation that she can quote in her thesis. When looking at the notes for the profile in Beyond 20/20, only the criteria for suppressing data about income is explained.
AnswerWe have received the following response from subject matter regarding your questions:
“Unfortunately, the DA suppression list is nowhere on our website…we do have the CSD suppression list in the notes for the data quality and confidentiality of the community profiles. I compiled a list of suppressed DAs (approximately 8,909 DAs) in a spreadsheet for you (please see attached spreadsheet). This suppression list is not very accessible for previous Censuses.
To answer your questions: In your example below (question 1) of DA 12090310, this DA in Halifax is suppressed. You can see from Geosuite 2001 that this DA does not meet the population threshold for the data quality and confidentiality rules.
For question 2, the specific DA 12090292 is not suppressed for 100% data but it is suppressed for 20% data (the non-response rate for the 20% was over or equal to 25%). You can find an explanation in the Beyond 20/20 notes in the data quality flag section.
For example, in the Beyond 20/20 file in the geography row you will see the following type of description for a particular DA. There is a 5 digit code containing the data quality flags which indicate whether suppression at the data quality level has been applied. See example below with the description of the DQ flags:
2006 Census – Data quality flags and descriptions
Data quality flags for place of residence
Digit Description
1st (0XXXX) Incomplete enumeration flag
2nd (X0XXX) 100% data quality flag
3rd (XX0XX) Population and dwelling counts error flag
4th (XXX0X) 20% sample data quality flag
5th (XXXX0) 2001 adjusted population flag
Flag description
Incomplete enumeration flag
0 = Default.
1 = Incompletely enumerated Indian reserve or Indian settlement (suppressed).
2 = Excludes census data for one or more incompletely enumerated Indian reserves or Indian settlements.
100% data quality flag
0 = Default.
1 = Data quality index showing, for the short census questionnaire (100% data), a global non-response rate higher than or equal to 5% but lower than 10%.
2 = Data quality index showing, for the short census questionnaire (100% data), a global non-response rate higher than or equal to 10% but lower than 25%.
3 = Data quality index showing, for the short census questionnaire (100% data), a global non-response rate higher than or equal to 25% (suppressed).
Population and dwelling counts error flag
0 = Default.
1 = An error exists in the 2006 population and dwelling counts for this area. For further details, please refer to the population and dwelling counts data section of the 'Notes' file.
2 = In 2001, the population and/or dwelling counts for this census subdivision were found to be incorrect. Since it is not possible to make changes to the 2001 Census data presented in these tables, the 2001 data should be used with caution. For further details, please refer to the population and dwelling counts data section of the 'Notes' file.
3 = Both the 2006 and 2001 population and/or dwelling counts for this area were found to be incorrect. Since it is not possible to make changes to the census data presented in these tables, these counts should be used with caution. For further details, please refer to the population and dwelling counts data section of the 'Notes' file.
20% sample data quality flag
0 = Default.
1 = Data quality index showing, for the long census questionnaire (20% sample data), a global non-response rate higher than or equal to 5% but lower than 10%.
2 = Data quality index showing, for the long census questionnaire (20% sample data), a global non-response rate higher than or equal to 10% but lower than 25%.
3 = Data quality index showing, for the long census questionnaire (20% sample data) or the short census questionnaire (100% data), a global non-response rate higher than or equal to 25% (suppressed).
2001 adjusted population flag
0 = Default.
1 = 2001 adjusted count; most of these are the result of boundary changes.
Note:
1. The 100% and 20% sample data quality flags do not apply to the population and dwelling counts.
Data quality flag for place of work
Digit Description
4th (XXX0X) 20% sample data quality flag
Flag description
20% sample data quality flag
3 = Data quality index showing, for the long census questionnaire (20% sample data) or the short census questionnaire (100% data), a global non-response rate higher than or equal to 25% (suppressed).
More detailed information can be found here: Data quality and confidentiality standards and guidelines (public): Introduction.”
