Stats NZ has a new website.

For new releases go to

As we transition to our new site, you'll still find some Stats NZ information here on this archive site.

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
Te Kupenga 2013 (English) – corrected
Embargoed until 10:45am  –  06 May 2014
Data quality

Survey objectives

The objectives of Te Kupenga were to:

  • measure engagement in te ao Māori (the Māori world) for (groups within) the Māori population, including traditional and modern ways of engaging
  • measure general well-being outcomes for (groups within) the Māori population
  • measure overall subjective well-being and whānau well-being for (groups within) the Māori population
  • allow analysis of how engagement in te ao Māori relates to the general well-being outcomes
  • allow analysis of the interrelationships between engagement in te ao Māori and general well-being outcomes, and subjective well-being and whānau well-being.


Target population

The target population for Te Kupenga was the usually resident Māori population of New Zealand, living in occupied private dwellings on 2013 Census night and aged 15 years or older.

The Māori population includes all individuals who identified with Māori ethnicity or Māori descent in the 2013 Census form.

The usually resident population excludes:

  • non-New Zealand diplomats and non-New Zealand members of their staff and households
  • members of non-New Zealand armed forces stationed in New Zealand and their dependants
  • overseas visitors who have been resident in New Zealand for less than 12 months and who do not intend to stay in New Zealand for a total of more than 12 months.

Occupied private dwellings are permanent or temporary dwellings that are occupied by a person or group of people and are not available for public use. The main purpose of a private dwelling is as a place of habitation, and it is usually built (or converted) to function as a self-contained housing unit. This includes:

  • separate houses
  • flats/units and apartments that are self-contained in respect of sleeping, cooking, dining, bathing, and toilet facilities
  • motor camps
  • tents
  • caravans
  • campervans
  • boats.

Survey population

The Te Kupenga survey population is the usually resident Māori population of New Zealand, staying in occupied private dwellings on the main islands of New Zealand (North, South, and Waiheke) on census night in 2013, and who were aged 15 years and over on census day (5 March 2013). Those who did not fill in an individual census form, or did not answer the ethnicity and descent questions, were excluded from the survey.

Survey methodology

Survey design

Te Kupenga was run as a post-census survey. A post-census survey provides a unique opportunity to run a large survey of a small sub-group of the population in a cost-effective manner. Administrative lists, such as the electoral roll, suffer from serious undercoverage; and using the Statistics NZ household survey frame is expensive because of the need to screen a large number of households in order to find the target population. Using the 2013 Census as a frame provided a degree of coverage not matched by any other single method.

In addition, information collected from respondents in Te Kupenga can be linked with their census responses to questions such as income and labour force status. This minimises respondent burden and helps to reduce data collection costs.

Sample design

The sample selection for Te Kupenga was conducted in four stages.

At the first stage, we selected a sample of primary sampling units (PSUs) from the Household Survey Frame (HSF). The HSF is the standard sampling frame Statistics NZ uses to select samples and manage overlap control for all of its household surveys. The HSF lists PSUs with attributes determined by data from the census. PSUs were then assigned to standard strata based on these attributes.

The second stage used the census to identify people within each stratum, from the PSUs that were chosen at the first stage. For Te Kupenga, there were three age strata – we selected a sample from within each age stratum.

The third stage of selection involved randomly subsampling one person from dwellings where more than one person was originally selected. So, if three people were selected after the second stage of selection, two were discarded at random. This ensured that only one person was interviewed from each private dwelling selected into the survey.

The final stage of selection consisted of the number of people interviewed in each PSU being capped at 23. This was achieved by randomly selecting 23 people from within all PSUs where more than 23 were selected at the third stage.

Reliability of survey estimates

Two types of error are possible in estimates based on a sample survey: sampling error and non-sampling error.

Non-sampling errors are all errors that are not quantifiable and include unintentional mistakes by respondents, variation in the respondent's and interviewer's interpretation of the questions asked, and errors in recording and coding data. We endeavour to minimise the impact of these errors by applying best survey practices and monitoring known indicators (eg non-response).

Sampling error can be measured. It quantifies the variability that occurs by chance because a sample rather than an entire population is surveyed. In Te Kupenga, sampling errors were estimated using a jackknife method. This method is based on the variation between estimates, and on taking 100 mutually exclusive subsamples from the whole sample. Sampling errors are quoted at the 95 percent confidence level. For example, if the estimated total number of people is 315,000, and the estimate is subject to a sampling error of plus or minus 7,250, or 2.3 percent (measured at the 95 percent confidence level), that shows there is a 95 percent chance that the true total number of people lies between 322,250 and 307,750.

Relative sampling errors (RSE) are the sample error as a percentage of the estimate. These figures do not give a precise measure of the sample error for a particular estimate, but do indicate its magnitude. Smaller estimates, such as the number of Māori who are unemployed, are subject to larger RSEs than larger estimates. Very small estimates are subject to such high sample errors (relative to the estimate) as to detract seriously from their value for most reasonable uses.

Imputation for item non-response

Imputation is the replacement of missing information with a best estimate of what the true value might be. We do this very selectively, to maintain the quality of the data. Imputation allows more questionnaire responses to be included in the final dataset. In Te Kupenga, imputation was only carried out on personal income, which was obtained from a respondent’s census information. Very little imputation was required, with 347 records being imputed for personal income.

Personal income was imputed using the donor method imputation. This involves matching the non-respondent (recipient) to a respondent (donor) for a particular question, based on a set of matching variables that are closely related to the missing variable. The method copies the missing information from the donor to the recipient. For example, if the personal income information is missing for a 35-year-old male, a male in the same age group is found as a donor of this information.

These were the only edits made to Te Kupenga data.

Response rate

The Te Kupenga questionnaire was answered by 5,549 individuals.

The target response rate for Te Kupenga was 75 percent. The achieved response rate was 74 percent. We calculated the response rate by dividing the weighted percentage of eligible individuals who responded to the survey by the estimated number of eligible individuals.

Collection methods

Respondents were offered the choice to complete the survey in either te reo Māori or English, with the layout of the questionnaire allowing respondents to switch from one to the other if necessary. Just 27 interviews were completed in either te reo Māori or a combination of te reo Māori and English, representing 0.5 percent of the total number of interviews.

We recruited 11 interviewers with Māori language understanding and ability for the collection of Te Kupenga, which gave respondents the opportunity to complete the survey in te reo Māori. They were able to undertake interviews in English too. We trained these collectors to interview using a computer-assisted laptop, as they were new to Statistics NZ. The 11 bilingual interviewers were located in the areas with a high proportion of Māori language speakers. If a respondent asked to complete the survey in Māori, a bilingual interviewer in the area closest to the respondent's location conducted the interview. At times, this required some bilingual interviewers to work outside their designated area.

Interviews were conducted using computer-assisted personal interviews and lasted an average of 40 minutes.

Reference period

The data collection for Te Kupenga took place from 4 June 2013 to 25 August 2013.

Change due to 22 February 2011 earthquake

Originally planned for 2011, Te Kupenga and the 2011 Census were both postponed following the 22 February earthquakes in Christchurch.

Consistency with other datasets

Use of the Māori language has been measured by different sources over the last 20 years. Since 1996, the population census has asked New Zealanders which languages they can have a conversation in about a lot of everyday things, of which Māori is one response option. Statistics NZ has also run two post-census surveys (in 2001 and 2013). The 2001 Survey on the Health of the Māori Language (HMLS) was sponsored by Te Puni Kōkiri. It collected more in-depth and wide-ranging information on Māori adults’ te reo Māori speaking ability and usage.

These data sources all use self-rated measures of a respondent’s te reo Māori ability. While international research has shown that self-rated measures are generally predictive of language ability, the context they are collected in can affect results. This context should be considered when comparing data sources.

Data from these sources may show differences for several reasons – differences in methodology, scope, coverage, non-response rates, data being collected at different periods of time, or differences in question wording and collection method. We advise data users to familiarise themselves with the strengths and limitations of individual data sources before making a comparison with Te Kupenga data.

The te reo Māori content in Te Kupenga used a selection of questions from the 2001 HMLS. We made this decision to increase comparability between the two surveys. Te Kupenga also used a similar methodology to the 2001 HMLS. However, there are other administrative differences between the two surveys that are likely to have influenced time-series results. In particular:

  • The 2001 HMLS was about Māori language specifically, whereas Te Kupenga is about Māori well-being more generally.
  • The 2001 HMLS predominantly used Māori interviewers (73 percent), whereas Te Kupenga used predominantly non-Māori interviewers.
  • For the 2001 HMLS, the majority of the interviewers were fluent te reo Māori speakers, so a fluent speaker was very likely to be available immediately, on the doorstep, to conduct the interview in Māori if requested by the respondent. In contrast, respondents requesting a te reo Māori interviewer in Te Kupenga had to re-book their interview time and day.
  • In the 2001 HMLS, 12 percent of interviews were conducted in te reo Māori or a combination of te reo Māori and English, compared with 0.5 percent in Te Kupenga.

In summary, the 2001 HMLS provided a stricter setting for determining speaking ability.

The 2001 HMLS surveyed individuals who identified with Māori ethnicity, whereas Te Kupenga surveyed individuals who identified with Māori ethnicity or Māori descent. To increase comparability between the two surveys, the te reo Māori analysis contained in this release is restricted to the Māori ethnic group only.

Note: the 2001 HMLS data was re-weighted, to place it on the same 2006 Census-based population estimates time series as Te Kupenga.

Statistics NZ will produce a topic-specific te reo Māori report following this information release. It will include a fuller exploration of the different data sources (censuses and surveys), time-series analysis, and methodology exploration.

Presentation of information


The survey has weights attached to allow the survey sample to be used to describe the whole Māori population resident in New Zealand. Weights are at the person level only.

Confidentiality and suppression

Estimates with very few contributors are deemed a risk to respondents’ confidentiality. Estimates with an estimated population of less than 500 are suppressed. This is flagged with an S in tables.

Rounding and percentages

All percentages shown in the text are calculated from weighted data and then rounded to the nearest whole number. The percentages in the tables are rounded to one decimal place. To improve the readability of the data, the calculation of percentages excludes residual categories (eg ‘don’t know’ and ‘refused’) in the population base from which percentages are calculated.

Calculating percentages

Unless otherwise stated, all percentages and ratios in this report exclude responses that cannot be classified (eg 'don’t know', 'refused').

Relative sampling errors

Estimates with high relative sample errors (RSE) are not reliable. Therefore all estimates with an RSE of 100 percent or greater are suppressed. Estimates with an RSE of 30 percent to 49.9 percent should be viewed with caution (flagged in tables by an asterisk*), and those with an RSE of 50 percent to 99.9 percent should be considered unreliable (flagged by**).

More information

See Te Kupenga for more information.


We would like to acknowledge the support of Te Puni Kōkiri in the development of Te Kupenga.


While all care and diligence has been used in processing, analysing, and extracting data and information in this publication, Statistics NZ gives no warranty it is error-free and will not be liable for any loss or damage suffered by the use directly, or indirectly, of the information in this publication.


Our information releases are delivered electronically by third parties. Delivery may be delayed by circumstances outside our control. Statistics NZ does not accept responsibility for any such delay.

Crown copyright©

Creative Commons logo.

This work is licensed under the Creative Commons Attribution 3.0 New Zealand licence. You are free to copy, distribute, and adapt the work, as long as you attribute the work to Statistics NZ and abide by the other licence terms. Please note you may not use any departmental or governmental emblem, logo, or coat of arms in any way that infringes any provision of the Flags, Emblems, and Names Protection Act 1981. Use the wording 'Statistics New Zealand' in your attribution, not the Statistics NZ logo.

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+