Appendix E – Sampling and Limitations

Sampling and Response Rates

There were 12,726 migrants who were eligible to receive a questionnaire during the surveying period. Of these, 11,004 had New Zealand addresses.

In total, 1,320 of the 11,004 eligible migrants with known New Zealand addresses were randomly selected to participate in the research – a sampling rate of 12 percent.

Of the 1,054 questionnaires that were successfully posted out (266 of the 1,320 questionnaires were returned for reason of wrong address), n=601 migrants responded to the survey. This equates to a 57 percent response rate for migrants with known addresses in New Zealand.

Limitations

The findings detailed in this report have four main limitations, which need to be considered in interpreting and generalising the results:

• The overall representation of those migrants who participated in the survey compared to the actual eligible migrant population.
• The length of time some migrants spent in New Zealand prior to having their residence approved.
• The response rate and non-response bias.
• The margin of error of ± 4.0 percent (at the 95 percent confidence level) associated with a randomly selected sample of n=601 migrants.

Representation of migrants

The survey results are comprised of a sample of migrants for whom the Department of Labour had address details. Table 14 identifies the differences between the eligible population of migrants (n=12,726) and the achieved sample of survey respondents (n=601).

As shown in the table, there is little variance between the percentages of principal and secondary applicants, as well as males and females, in the achieved sample of respondents when compared with the eligible population.

Migrants who were approved onshore, however, are significantly over-represented among the survey respondents (83 percent compared with 65 percent of the eligible population), while those approved offshore are significantly under-represented (17 percent compared with 35 percent of the eligible population).

Migrants approved through the SMC are under-represented among the survey respondents, while those approved under the GSC and LTBV categories are over-represented. Similarly, migrants from Great Britain and South Africa are over-represented, while those from China, the Philippines and Malaysia are under-represented.

Those aged 20 to 29 years old are under-represented among the survey respondents, while those aged 40 years and over are over-represented, when compared with the total population of migrants who were eligible to participate during the study period of January 2007 to June 2007.

Length of time in New Zealand

The purpose of the SEFS is to collect migrant settlement outcomes and experiences 12 months on. Table 29 shows that the majority of the survey respondents (83 percent compared with 65 percent of the eligible survey population) were already living in New Zealand when they were approved for residence. Their responses are over-represented compared to migrants approved offshore (17 percent of the survey sample).

Therefore, the greater proportion of onshore approvals in the achieved survey sample should be taken into account when interpreting the results from the survey (migrants approved onshore would have an increased familiarity with living in New Zealand than those approved offshore). The fact that those approved onshore would have been already living in New Zealand for varying lengths of time should also be taken into account.

Response and non-response bias

As discussed above, there were 12,726 migrants who were approved for residence through the Skilled/Business Stream in New Zealand between January 2006 and June 2006. The Department of Labour had address details for 11,004 of these individuals, of whom a sample of 1,320 respondents were posted survey questionnaires. Of these, 266 were returned to New Zealand as having incorrect addresses (a 20 percent wrong address rate), and 108 individuals refused to participate in the survey when contacted by telephone as part of the reminder process.

Of the 1,054 potential respondents with correct address details, 601 responded to the survey – a 57 percent response rate. However, as the degree to which non-respondents differ from respondents in relation to their settlement experiences is not known, care should be taken when interpreting and extrapolating the survey results to the entire eligible population of migrants who took up residence in 2006, or in previous years.

It should also be noted that the survey was administered in English. As such, it may not represent the experiences and views of secondary migrants with poor or little written English language skills, although this is not an issue for principal applicants as they are tested for English as part of the application process.

Despite these limitations, the survey findings covered in this report provide a useful body of information on the short term settlement experiences and outcomes of Skilled/Business Stream migrants, and how these change over time.