|
|
|
|
A total of 6,546 respondents were interviewed for the survey throughout Scotland between February and June 1999 by interviewers from System Three and MORI Scotland. The sample was drawn from the Small User File of the Postcode Address File (PAF), a listing of all active address points maintained by the Post Office. The survey interview, which lasts an average of 41 minutes, is carried out in respondents' homes by interviewers using Computer Aided Personal Interviewing (CAPI).
The interview has a two-part structure. The first part is carried out with the Highest Income Holder (HIH), or his/her spouse or partner, and collects mainly factual information about the composition and characteristics of the household. Some questions are asked in respect of each household member.
The second part of the interview takes place with an adult member of the household selected at random and focuses more on individual attitudes and behaviours. (In many cases, of course, both parts of the interview are completed by the same person).
The response rate for the first part of the survey (taking account of 'deadwood' in the sample) was 72%. Among households that participated, the response rate for the second part of the survey was 94%, and interviews were achieved with 6,150 adults.
The data have been weighted to take account of the unequal probabilities of selection inherent in the sample design due to (a) the over-sampling of smaller local authority areas relative to their household population in order to obtain a minimum number of interviews in each area and (b) the under-sampling of adults resident in multi-person households because only one 'random adult' is interviewed in each household.
Comparisons with external sources, such as Census data and the 1996 Scottish House Condition Survey, suggest that the weighted data are broadly representative of the Scottish household population; consequently, no further weighting has been carried out.
In many instances, the survey uses the same questions as in other government surveys in order to facilitate comparisons on a consistent basis. Due to differences in sampling and survey methods, however, the results from this survey will inevitably be slightly different from other government surveys.
|
|
|