CAM Data Reliability
Last updated:
April 2008
Response Rates
The Accommodation Survey aims for 100% coverage of mainstream commercial accommodation providers (that is, a census).
In practice an overall response rate of between 76 and 80% is usually achieved. The response rate varies amongst the different accommodation types as show in the following table.
Generally the data quality of the survey is considered to be good. The main weakness is in the hosted accommodation type as many small establishments are not covered by the survey.
Response rates by accommodation type for January 2008
|
Accommodation Type |
% of establishments providing response for guest nights, guest arrivals & stay unit nights |
% of guest night estimate from establishments that provided guest night response |
% of origin of guest estimate from establishments that provided origin of guest response |
| Hotels |
79 |
86 |
64 |
| Motels |
78 |
80 |
64 |
| Hosted |
83 |
83 |
71 |
| Backpackers/hostels |
77 |
78 |
55 |
| Caravan parks/camp grounds |
83 |
89 |
73 |
| Total |
79 |
85 |
67 |
Imputation / Missing Data
Imputed values are used where data is missing, based on the characteristics of similar establishments in the same or similar regions. Imputation introduces unknown errors into the estimates, and this should be borne in mind by users of the data. The size of these unknown errors is difficult to quantify.
Non-Sampling Error
Other non-sampling errors occur for reasons such as respondent error, frame quality and errors in processing. While every effort is made to minimise these types of error they will still occur. It is not possible to quantify their effect.
Also, there is under coverage of small accommodation providers that are outside the scope of the survey (see Survey Coverage).
Source: Statistics New Zealand
ˆ Back to top