The process used to determine content for the 2011 Census is described below.
Phase 1 – Internal and initial external consultation
During this phase, census staff worked closely with Statistics NZ topic experts to develop an initial position within the context of the Programme of Official Social Statistics and its associated surveys. Meetings were held with key policy agencies and expert data users. This consultation led to the publication of the 2011 Census Content: Consultation guide.
Phase 2 – Public consultation and evaluation of submissions
The public consultation phase began in late May 2008 with the publication of the consultation guide. Public consultation closed in July 2008. Some consultation with interest groups, ethnic groups, government agencies, and expert users also took place following the release of the consultation document. An evaluation of submissions from the public consultation phase was published in the 2011 Census Content: Submissions report in November 2008.
At the close of the public consultation phase, all submissions were summarised and evaluated by Statistics NZ's topic experts. Evaluations were carried out within the context of all known information on a topic at that stage and recommendations for further action were made. To ensure a consistent approach to evaluation, a group of moderators independently assessed the recommendations for further action against the criteria detailed in the 2011 Census Content: Consultation guide.
Submitters were generally supportive of Statistics NZ's minimal content change strategy for the 2011 Census, recognising the need to maintain consistency between censuses while reflecting changing data needs over time.
Phase 3 – Questionnaire development, testing, and classification reviews
As a result of the evaluation process, some topics were moved into the next phase of the content development process. This involved assessing the value and cost of implementing any change, analysing the effectiveness of proposed questionnaire changes through cognitive testing, and evaluating the results from the March 2009 Census Test.
In order to maintain relevance, the outcomes of Statistics NZ classification and standard reviews and their impact on the census have also been considered. Where updated classifications and standards have the potential to improve data quality, and can be incorporated into existing processing systems, they will be used for the 2011 Census. A key consideration when determining census content is the impact any changes will have on the subsequent parts of the census process, particularly the cost of altering collection and processing systems.
Most importantly, all changes have been assessed in terms of their effectiveness in improving the quality of the data collected for that topic. The outcomes of this phase of content development are outlined in this report.
Quality management and 2011 Census content development
Census topics and variables are ranked in terms of three 'quality levels' based on uses of the data and constraints of the census process. The three levels – foremost, defining, and supplementary – are used to guide delivery of quality data and products throughout the census life cycle. Their use ensures census data is fit-for-use, and assists in the allocation of time and money for reaching data quality standards for particular variables.
The three quality levels are described more fully below.
Foremost topics/variables are core census variables and their outputs are the key reasons for conducting a census. They need to be included in the census to meet the requirements of the Statistics Act 1975 and the Electoral Act 1993. The outputs for these variables are used for maintaining the accuracy of population estimates. They are given the highest priority in terms of quality, time, and resources across all phases of the census.
Defining topics/variables define key subject populations that the census provides measures for. These are important for policy development, evaluation, and monitoring. Defining variables are used frequently in cross-tabulations with foremost variables. They represent key sub-populations and measures that are of high public interest. These variables are closely linked to the main purpose of a census of population and dwellings, and in the New Zealand context, data from them may only be available in detail, for example at subnational level, from the census. These variables have second priority in terms of quality, time, and resources across all phases of the census.
Supplementary topics/variables do not fit directly with the primary purpose of a census of population and dwellings, but are important to certain groups. These variables have third priority in terms of effort and resources. However, there are minimum quality standards to be met in order to make the output data fit-for-use.
The focus on data quality for the 2011 Census has been a key component in determining the topics that were the basis of the consultation and submissions process in 2008. The three quality levels have also informed the testing and evaluation of proposed changes that have occurred since the close of the submissions process. The results of this stage of content development are contained in the following section.