« Previous | Contents | Next »
Listen
Annex D: Methodology
Conduct of the research
1.5 MORI Scotland and TNS Social jointly conducted the 2003 and 2004 surveys. MVA conducted the 1993, 1996 and 2000 surveys.
Response rates
1.7 The 2004 face-to-face SCVS interviewed 5,007 adults aged 16+ between July and November 2004 and achieved a response rate of 67 per cent. Of these, 4,424 undertook the self-completion questionnaire. The response rate for self-completion element of the 2004 survey was therefore 59 per cent (based on all of the eligible sample).
Sampling and weighting
1.8 Like previous sweeps, the 2004 SCVS used a probability sample which selected households at random from the Postcode Address File and according to a multi-stage stratified design. The sample points were clustered, but were spread across the whole of Scotland including the Highlands and larger islands. Once selected households were contacted, one respondent was selected at random (using the Kish grid method) from all adults aged 16+ who were resident at the address. No substitution of households or individual respondents was permitted.
1.9 Weighting was conducted on the main survey data to compensate for the design elements of the survey (elements that altered the probability of sampling units being from a simple random sample, such as weights to compensate for differential probability of selection according to the number of adults in the household) and to compensate for non-response bias where this was observable by comparing the survey data with population estimates. The self-completion element was subject to the same selection probabilities and biases as the main survey, with additional scope for bias arising from non-response to the self-completion element. The main weight used for the self-completion element was therefore the individual weight used for the main survey data. However, small additional weighting was undertaken to adjust for non-response to the self-completion element (calculated by comparing the weighted profile of adults who completed a questionnaire with that of the adult population).
Self-completion questionnaire
1.10 The 2004 self-completion questionnaire is included in Annex E. The wording of the questions was the same as previous waves, but some minor amendments were made in order to improve the layout and make the form more user-friendly.
1.11 Respondents were asked to fill in the self-completion questionnaire at the end of the main face-to-face interview. Interviewers handed the respondent the questionnaire along with a self-sealing envelope. They emphasised the confidentiality of the survey and asked the respondent to read the front page carefully. Respondents then completed the questionnaire and returned it to the interviewer in the sealed envelope. Interviewers could assist the respondent if required, but not with the question wording or answers. The questionnaire took 5-10 minutes to complete.
Use of the bogus drug
1.18 Including the name of a fictitious drug is a technique that is commonly used in drugs studies in order to deal with the problem of respondents exaggerating their drug use - for reasons of 'bravado' or any other reason. 16 When this technique was first used, the standard practice was to exclude from further analysis any respondents who claimed to have heard of, or used, the bogus drug. This was done on the assumption that they must be lying and their other answers would be unreliable. More recently, however, there has been some debate as to whether this approach is appropriate. It is argued that some people may report use of a fictitious drug because they know they have tried almost everything they have been offered, so they genuinely think that they 'must' have used the bogus drug too. Excluding these respondents therefore introduces sample bias by removing some of the multiple drug users (of whom there are relatively few). 17 It has also been argued that some respondents may actually have heard of the bogus drug, due to the increase in social surveys carried out in recent years which refer to the same bogus drug 18. However, it seems unlikely that this would affect more than a very small proportion of respondents.
1.20 There were 20 respondents in the 2004 survey who claimed to have used the bogus drug. These 20 have been excluded from the analysis.
Further details on methodology
1.12 Full details on the methodology for the 2004 SCVS are contained in the technical report, which is available from the Scottish Executive on request.
« Previous | Contents | Next »