Changes to sample design/process

This section gives an overview of how we changed the sampling process in each year we did the New Zealand Crime and Safety Survey (NZCASS).

You can find more information in the NZCASS Technical manual [PDF, 5.3 MB]

Changes in 2009

There were four main changes to the sampling process in 2009.

Primary sampling unit (PSU) used

Before 2009, the PSUs used for sampling were Nielsen area units (NAUs). In 2009, a new fieldwork provider was contracted to undertake the NZCASS. The new fieldwork provider replaced NAUs with Statistics NZ’s meshblocks (MBs).

Number of PSUs selected

The number of PSUs selected changed from 800 NAUs selected for the main sample and 320 separate NAUs for the Māori booster sample in 2006, to 1000 MBs in total in 2009.

Selection of Māori booster areas

In 2006, the areas for the Māori booster sample were selected independently from the main sample areas, meaning that only Māori were eligible. In 2009, the same MBs were selected for both the main and the Māori booster samples.

Unstratified sample

An unstratified sample of PSUs was selected in 2009, in contrast to a stratified sample in 2006.

Back to top

Changes in 2014

The two main changes to the sampling process in 2014 aimed to:

  • improve efficiencies in contacting Māori respondents
  • achieve a higher response rate for both the main sample and the Māori booster sample
  • reduce fieldwork costs.

The impact of these changes (if any) is discussed in the NZCASS Technical Manual.

NZ postal address file (PAF)

In 2014, the NZ Postal address file (PAF) was used to sample households as part of the sampling process. Enumeration was conducted to ensure that any households not listed on the PAF were captured and incorporated into the process.

The fieldwork provider used the PAF to send out pre-survey letters to selected households. This letter gave householders a level of comfort when a stranger called at their door. The letter also gave households the chance to make an appointment with the interviewer; this encouraged participation.

Māori electoral roll

In 2014, we used the Māori electoral roll as part of the sampling process. This helped increase the likelihood of approaching households with Māori respondents to be part of the Māori booster sample.

We got approval to use this roll information because the NZCASS is a social science survey and collects a range of information, including data on the physical and emotional effects of crime and victimisation.

Back to top

This page was last updated: