Accessibility Home page Site Map Search Contact Us Skip to Content Govt.nz
Logo for Print
   

DTS Data Reliability

Last updated: May 2008

The Domestic Travel Survey (DTS) draws on a sample of between 15,000 New Zealand residents per year to represent the behaviour of all domestic travellers in New Zealand.

 

The results for the DTS are subject to measurement errors, including both sampling and non sampling errors. These errors should be considered when analysing the results from the survey.

 

Sampling Errors

Sampling errors arise from estimating a population characteristic by looking at only a sample of the population rather than the entire population. If refers to the difference between the estimate derived from a sample survey and the “true” value that would result if everyone in the population was surveyed.

 

The DTS sample includes 15,000 New Zealand residents out of the 4 million New Zealand residents each year.

 

The sampling error:

  • Is different for each estimate in the survey
  • Generally decreases as the sample size increase
  • Depends on the size of the population
  • Depends on the variability of the characteristics of interest in the population.

 

The total expenditure in the year ended December 2008 was $8.1 billion. This estimate has a sampling error of plus or minus 2.6% (at the 95% confidence level). This means that there is a 95% chance that the true expenditure estimate lies between $7.8 and $8.3 billion.

 

For the DTS, sampling errors of selected estimate have been modelled as a function of the estimate itself.


 

DTS Generalised Sampling Errors (at the 95% confidence level)

Size of Estimate

Day Trips

Overnight Trips

All Trips

Nights

Expenditure

1,000

>100%

>100%

>100%

>100%

>100%

2,000

>100%

>100%

>100%

>100%

>100%

3,000

>100%

>100%

>100%

>100%

>100%

4,000

>100%

>100%

>100%

>100%

>100%

5,000

>100%

>100%

>100%

>100%

>100%

10,000

>100%

>100%

>100%

>100%

>100%

20,000

>100%

>50%

>100%

>100%

>100%

30,000

>100%

>50%

>100%

>100%

>100%

40,000

>100%

>50%

>100%

>100%

>100%

50,000

>100%

>50%

>100%

>100%

>100%

100,000

>50%

40.4%

>50%

>50%

>100%

200,000

>50%

27.8%

>50%

>50%

>100%

300,000

46.8%

22.4%

43.1%

46.1%

>100%

400,000

40.4%

19.2%

36.9%

40.3%

>100%

500,000

36.1%

17.0%

32.7%

36.3%

>100%

1,000,000

25.3%

11.7%

22.5%

26.2%

>100%

2,000,000

17.8%

8.1%

15.5%

18.9%

>100%

3,000,000

14.4%

6.5%

12.5%

15.7%

>100%

4,000,000

12.5%

5.6%

10.7%

13.7%

>100%

5,000,000

11.1%

4.9%

9.5%

12.3%

>100%

10,000,000

7.8%

3.4%

6.5%

8.9%

>50%

15,000,000

6.3%

2.7%

5.2%

7.4%

>50%

20,000,000

5.5%

4.5%

6.4%

>50%

30,000,000

3.6%

5.3%

>50%

40,000,000

3.1%

4.7%

44.2%

50,000,000

39.5%

100,000,000

27.9%

200,000,000

19.7%

400,000,000

14.0%

600,000,000

11.4%

800,000,000

9.9%

1,000,000,000

8.8%

5,000,000,000

3.9%

7,000,000,000

3.3%

 

Non Sampling Errors

Aside from the sampling error associated with the process of selecting a sample, a survey is also subject to a wide variety of other errors. These errors are commonly referred to as non sampling errors. Non sampling errors are defined as errors arsing during the course of all survey activities other than sampling. Unlike sampling errors they are generally difficult to measure.

 

Some examples of non sampling errors include:

  • Respondent errors
  • Poor questionnaire design
  • Non response

 

These types of errors are minimised through the use of best survey practices and monitoring.

 

ˆ Back to top