« Previous | Contents | Next »
Listen
Chapter 1 Introduction
Background
1.1 This report presents the findings on self-reported drug use, and attitudes towards drugs and drug-taking, from the 2004 face-to-face Scottish Crime and Victimisation Survey ( SCVS) 3.
1.2 The SCVS focuses primarily on experiences of personal and household crime, and is designed to provide a measure of crime complementary to police recorded crime statistics. However, there is an additional self-completion questionnaire which asks adults about their drug misuse and experiences of domestic abuse. Questions on these topics are particularly sensitive and the self-completion mode is considered more effective in eliciting truthful answers.
1.3 The self-completion module has been a feature of the Scottish Crime Survey since 1993 and has remained largely unchanged - enabling a comparison of trends over time. Detailed results from previous sweeps (in 1993, 1996, 2000 and 2003) have previously been published 4.
1.4 Details of the methodology used in the 2004 SCVS are contained in Annex D.
Limitations of the data
Limitations of self-report drug surveys
1.5 Fraser noted the following limitations of self-report drug surveys in the following discussion of the 2000 findings - these apply equally to the other sweeps of the survey:
"Although self-report drug surveys are a very useful method of monitoring trends, it is important to note their limitations. Firstly, more serious or 'chaotic' drug users are likely to be under-represented as they are more likely to be homeless, in prison (or another institution), or simply never available for interview. Separate research would be required to reach such groups. Secondly, given the illegal nature of much drug misuse, it is likely that there is an element of under-reporting within the survey, particularly with regard to the use of more heavily stigmatised substances (such as heroin and crack cocaine). The use of these 'harder' drugs would be better explored through other methods 5. Thirdly, it is also possible that people will forget occasional use of drugs, particularly if this was a long time ago, or if they regard themselves as 'non-users'. Finally, it should be borne in mind that raw numbers used in the report are relatively small. For example, only 130 of the respondents [in the 2000 survey 6] were aged 16 to 19. These factors together suggest that the survey will, to some degree, underestimate the actual level of drug misuse in Scotland. Some over-reporting may also occur, and it is for this reason that [a] bogus drug is included in the list of controlled substances. Previous surveys have found that there is very little over-reporting of this kind and the same appears to be true in 2000 with only 18 respondents reporting the use of [the bogus drug] and these may have been mistakes rather than fabrications." 7.
Non-response bias in the sample
1.6 In addition to the under-representation of more serious and 'chaotic' drug users discussed above, there is also the potential for non-response bias stemming from the fact that, in any random pre-selected survey, people who are out more often are less available for interview and are likely to be under-represented. This has the potential to lead to bias where the subject being investigated is correlated with being out more often. Drug-use is one such subject 8. In particular, this is likely to lead to some under-representation of younger, recreational drug users in comparison with younger people who do not use drugs. In other words, the young people who are at home more often, and are more likely to be interviewed, are less likely to use drugs than the young people who are out more often, and are less likely to be interviewed.
1.7 While corrective weighting has been used to correct observable bias in the sample ( e.g. to correct for the under or over representation of particular age-groups) it cannot correct for differences in the drug-taking behaviour of those who are interviewed and those who are not.
Results for the 20-24 age group
1.8 In the 2004 survey, there was a relatively poor response rate from respondents aged 20-24: there were 195 achieved interviews with respondents in this age-group in 2004 compared with 257 in 2003. A greater weight has therefore had to be applied to the 20-24 year olds in 2004 than was the case in 2003. The lower response rate from this group in 2004 increases the likelihood of the non-response bias discussed above, and there is an indication of this in the lower levels of drug use reported by this group compared with 2003 9. There is, of course, no way of measuring the extent to which the apparent difference reflects a real change in behaviour, non-response bias, or an increase in the under-reporting of drug misuse. Where results for different age groups are reported, the results for the sub-group 20-24 should therefore be treated with particular caution.
1.9 There will, of course, be a degree of non-response bias in all sweeps of the survey, which is likely to under-estimate levels of drug-use. This is likely to be more highly associated with younger age groups.
A note on the interpretation of the results
1.10 Unless otherwise stated, all data presented in this report has been weighted. The unweighted base sizes (in the form 'n=') have been provided beneath all tables and figures.
1.11 Only statistically significant differences (between sub-groups or different sweeps of the survey) are commented on in the text. This is true at the 95 per cent confidence limit. In other words, we can be 95 per cent sure that the difference is not due to chance. Statistical tests such as 2-tailed t-tests were carried out where appropriate.
1.12 Within the main body of the report, most results have been rounded to whole numbers, except on some occasions where results of between 0 and 1 per cent have been rounded to one decimal place. In the tables in the Annexes, results have been rounded to one decimal place.
1.13 Although the self-completion questionnaire was completed by all respondents to the SCVS (including those aged 60+), to facilitate comparisons with previous waves and with the British Crime Survey ( BCS), all results reported here are based on respondents aged 16-59 only. 'All' (in base text etc.) refers to all aged 16-59.
Structure of the report
1.14 The report is structured as follows. Chapter 2 focuses on the prevalence of drug use. It also examines trends over time and the characteristics of drug users. A brief comparison is made with results for England & Wales from the BCS. Chapter 3 provides details of the types of drugs used and methods of drug taking. Chapter 4 looks at those who were offered drugs and the types of drugs offered. Chapter 5 examines awareness of drugs, attitudes towards drug taking and perceptions of harm.
« Previous | Contents | Next »