Home → Publications → Reports → Research and public policy series → 67 → Technical appendix (in: Alcohol, drugs and crime : a study of juveniles in detention)
Alcohol, drugs and crime : a study of juveniles in detention
Jeremy Prichard and Jason Payne
ISBN 0 642 53891 3 ; ISSN 1326-6004
Canberra: Australian Institute of Criminology: 2005
(Research and public policy series, no. 67)
Download
- Technical appendix (PDF 257kB)
- Full report (PDF 1.4MB)
Technical appendix
- Methodology
- Reliability of estimates
- Weighting
- Limitations of the study
- Measuring drug and alcohol dependency
The data presented in this report are based on interviewer-administered questionnaires with juvenile detainees. Participation in this study was voluntary and informed consent was obtained. Two jurisdictions stipulated that parental consent also be obtained. The interviews were conducted between December 2003 and December 2004 in all states and territories.
Methodology
Sampling
In comparison with adult male prisons, Australian juvenile detention centres accommodate very small numbers of young people. For this reason, a census of the detainees was attempted in all jurisdictions. In addition to youths who had been sentenced to detention (n=200), juveniles on long-term remand participated (n=171). Several young people were excluded from participating because: they were incapable of giving informed consent; they were identified as being emotionally or psychologically vulnerable to participation; they represented a potential danger to themselves or others; or they were non-English speaking.
Content of the questionnaire
The main sections of the questionnaire asked the juveniles about their:
- socio-demographic background (Indigenous status, age, education, parental status and living arrangements);
- home environment, including family substance abuse, their history of abuse and neglect and current levels of contact with significant others;
- mental and emotional wellbeing;
- current charges;
- previous and current periods of detention;
- criminal history for 10 offence categories;
- patterns of use for seven categories of substances;
- treatments received for substance use; and
- perceptions of the criminal justice system.
Interview procedure
In the days preceding the data collection at each detention centre, announcements were made to the juveniles about the general nature of the research project and the voluntary basis of participation. Senior detention centre staff determined which youths would be eligible to participate, giving consideration to age, capacity to consent, psychological and emotional wellbeing, prior history of physical violence, and their English language abilities. In the two jurisdictions where parental consent was required, one refusal from parents caused the youth to be designated as ineligible for participation.
Detention centre staff approached the eligible detainees and informed them that a researcher would like to conduct an interview. If the youths indicated that they wanted to know more about the interview, they were accompanied to the interviewer who read a standard statement seeking the juveniles' informed consent. The staff remained within close proximity to the interview room without invading on privacy. Interviewers were also provided with personal alarm systems for their safety.
At set junctures of the interview, the participants were reminded that they could terminate the interview at any stage. They were also asked to reconfirm their consent periodically. Detention centre staff were not allowed to read the completed questionnaires, nor were they able to peruse a blank questionnaire. All materials were kept in briefcases or other containers while interviewers were on detention centre premises. The containers were removed at the conclusion of the daily data collection period. Over the duration of data collection in each jurisdiction, the completed questionnaires were stored in a lockable cabinet in the possession of the chief data collector. As the interviews were concluded in each state and territory the questionnaires were posted to the AIC, where the data entry process was overseen.
Informed consent
Given the explicit and sensitive nature of the information collected in this study, participants were afforded a measure of protection that meant they were not required to provide written (signed) consent. Instead, prior to the commencement of the interview, a descriptive statement was read to each participant that explained the scope and nature of the study and required offenders to provide verbal consent to participate. The introductory statement was designed in such a way to ensure that all participants understood that:
- the information collected from the interview would be held in the strictest of confidence;
- the interviewer may be required to breach confidentiality if the participants gave details of abuse they had suffered or serious offences they had committed for which they had not been charged;
- their participation was voluntary;
- they could not be individually identified in any published material;
- they could choose not to answer any question; and
- the interview could be terminated at any time.
Confidentiality
To guarantee confidentiality, a number of steps were taken so that neither the AIC nor the jurisdictional corrective services agency could identify individual participants and their responses. To ensure this, the contracted data collector applied a unique identifier to questionnaires. Then, as part of the data processing conducted at the AIC, another unique identifier was applied to each participant so as to further disguise identification.
Reliability of estimates
The response rate for the completion of the survey was 87 per cent (see Table A.1). Interviewers across Australia approached a total of 534 youths and, of these, 467 agreed to participate. The Northern Territory and the Australian Capital Territory had the highest response rates (100%) and Western Australia had the lowest (74%). Response rates in Western Australia were affected by difficulties associated with locating parents or guardians to provide consent. Numbers in the Victorian centres were lower than expected. Detention centre staff suggested this may have been partly caused by the fact that the Victorian interviews were conducted in the days prior to Christmas holidays when court referrals lessen.
A total of 95 participants were young adults, aged 18 or over. One participant from New South Wales was excluded because he was nine years old.
Table A.2 compares the DUCO juveniles sample with the participating jurisdictions.
The participants in the present study are representative of Australian juvenile detainees in terms of their age brackets: 13 per cent of the DUCO sample was aged 14 or less, and the national average figure as at 30 May 2003 was 14 per cent. Overall, the sample was also representative in respect to the sex of the detainees; females accounted for seven per cent of the DUCO sample and nine per cent of national detainees. However, the DUCO sample encompassed a higher proportion of Indigenous youths than were recorded nationally in June 2003, with rates at 59 per cent and 53 per cent respectively.
Weighting
As a result of over-sampling Indigenous juvenile detainees, it was possible that the drug use and offending comparisons between Indigenous and non-Indigenous youths were biased and did not accurately represent the general juvenile detainee population estimates. To examine this issue, the DUCO juvenile data were weighted so that the relative estimates were equivalent to the proportions seen in the national census of juveniles in detention (Charlton and McCall 2004). The data were weighted for age, Indigenous status and jurisdiction.
The differences seen between the Indigenous and non-Indigenous juveniles in this study held true after the application of the population probability weightings. This indicates that although over-sampled, the differences seen in this report are likely to hold true for the juvenile detainee population as a whole.
Given that the DUCO juvenile study was not a national census of juveniles in detention, the results presented in this report are unweighted.
Limitations of the study
A core component of the DUCO juveniles questionnaire required youths to provide sensitive information on the extent of their offending and illegal drug use, as well as their personal experiences of neglect, physical and emotional abuse. Much of the information provided by participants is in relation to activities that may not have been detected by law enforcement agencies, or disclosed to anyone else prior to the interview. Given the nature and content of this information, the DUCO project, like all projects of this nature, is limited to the extent to which the self-reported information is reliable.
In the field of criminology, research has shown that self-reported offending by adult prisoners is generally reliable, and that self-reported criminal histories are consistent with official records (Peterson et al. 1980). With respect to the accuracy of disclosures of illegal drug use, the DUMA project has established that a high degree of consistency exists between drug use reported in interviews and the detection of drug use in urinalysis (Milner et al. 2004), and that detainees with higher socio-economic status were most likely to underreport drug use (McGregor & Makkai 2003). However, findings are more erratic in regards to juveniles. Comparing self-reported substance use with urinalysis in Canadian adolescents, Williams and Nowatzki (2005) found that a quarter of juveniles underreported substance abuse. Strangely, one third overreported - that is, they self-report using substances but their urinalysis was negative. Williams and Nowatzki (2005: 299) concluded that self-reported substance use amongst juvenile samples appears to have only 'fair validity'.
These findings suggest that the potential limitations of self-report studies listed in the adult DUCO reports (Makkai & Payne 2003; Johnson 2004) may be of greater importance to the present study. The potential limitations referred to in the adult DUCO studies related to accuracy of memory recall, willingness to report sensitive or private experiences, and anticipated benefits of participation in terms of early release or other privileges.
There are other issues that need to be recognised that are germane in the juvenile setting. Commentators have pointed to the deficiencies that young people in the justice system may have with oral communication, such as formulating a 'coherent account of events' relating to their own offending (Snow & Powell 2004: 224). A related issue is the temptation to exaggerate substance use or criminal behaviour, apparently to impress or to shock interviewers, or perhaps with the general hope of standing out as special. Interviewers in DUCO noted that youths occasionally stressed their own excesses with pride (see further Ogilvie & Lynch 2001). Finally, it is true that juveniles have the benefit of recalling events that have generally occurred recently compared with the events recalled by adult offenders. On the other hand, commentators have underscored some of the difficulties caused by immature cognitive, moral and social development when young people analyse their own criminal behaviour (Prichard & Burton-Smith 2004). This may mean juveniles' constructs of, for example, assault may not be in accordance with legal definitions of that crime, causing over-estimation in some cases and under-estimation in others.
Measuring drug and alcohol dependency
The juvenile participants were not assessed clinically in terms of alcohol or drug dependency. The juvenile survey included the same dependency scale used in adult female DUCO (Johnson 2004). That scale presents six questions separately for alcohol and drugs. The questions are as follows.
In the six months before being arrested, did:
- you spend more time drinking alcohol/using drugs than you intended;
- you NOT do things you should have because of drinking or using drugs;
- you want to cut down your drinking or drug use;
- anyone complain or worry about your use of alcohol/drugs;
- you often find yourself thinking about drinking/using drugs; or
- you drink/use drugs because you were bored, angry or to make yourself feel better.
Table A.3 lists the youths' responses to these questions. The questions on drugs drew positive responses in 49-67 per cent of cases. Positive responses to the questions on alcohol were less frequent, ranging from 28 per cent to 54 per cent.
The six-item scale was developed by Hoffman et al. (2003) with a cohort of incarcerated adult males and females. According to Hoffman et al. (2003), positive answers to three or more of the items is indicative of dependence on alcohol or drugs.
Basic tests confirmed that the items were internally consistent within the adult female DUCO sample (Johnson 2004). Similar results were obtained with the juveniles. Factor analysis suggested that all items loaded on one factor and the factor accounted for 50 per cent of variance in the alcohol items and 51 per cent of the items relating to drugs. All six items correlated well for alcohol (Cronbach's alpha of .79) and drugs (.81). The results indicate that the questions used in the juvenile survey are measuring a common construct. Whether this construct is actually substance dependence will be discussed shortly. According to Hoffman et al.'s (2003) scale:
- 62 per cent of the DUCO youths meet the criteria for dependency on one or more substances;
- 48 per cent of youths meet criteria for alcohol dependency, and 66 per cent for drug dependency; and
- 81 juveniles (22%) meet the criteria for both alcohol and drugs.
There are reasons to be concerned about the applicability of Hoffman et al.'s (2003) scale for the juvenile population. Firstly, the scale was developed with the participation of 310 male and female adult prisoners and has not been tested upon a juvenile sample. Secondly, it is not clear how suitable some of the questions are for adolescents. For instance, while it may be significant for someone to complain to an adult about their use of alcohol or drugs, parents and guardians frequently complain to their teenage children about many aspects of their lives. Similarly, responsibilities that a juvenile failed to uphold because of alcohol or drugs might include weekend homework or household chores. Responsibilities of an adult, on the other hand, are more likely to include such things as paying bills, grocery shopping or going to work.
These issues could be overlooked, had previous empirical research produced clear estimates of the rates of dependency or addiction amongst juveniles (either in the general population or within detention centres). However, unlike the area of addiction amongst adults, no such evidence exists to date on adolescent dependence. A number of studies have assessed samples of young people (some being adults) using various instruments based on the Diagnostic and Statistical Manual (DSM), versions III and IV:
- from a sample of 1829 American juvenile detainees, almost half met DSM III-R criteria for substance abuse disorders in relation to one or more substances (McClelland et al. 2004);
- Lennings and Pritchard (1998) used the Alcohol Use Disorder Identification Test (AUDIT), which is designed to detect hazardous alcohol use. Seventy-one per cent of the 118 Queensland juvenile detainees met criteria for hazardous alcohol use;
- in Australia, of 1601 young adults in the general population (20 to 21 years) seven per cent met DSM-IV criteria for cannabis dependence (Degenhardt et al. 2002; Coffey et al. 2002); and
- of 2000 young Australians in the general population (20 to 21 years), 4.7 per cent met DSM-IV criteria for alcohol dependence (Bonomo et al. 2004).
Importantly, DSM-based instruments are not designed to detect dependence alone. Rather, their focus is more broadly upon disordered behaviour, which incorporates a number of constructs. For example, AUDIT measures frequency of alcohol consumption, dependency and social problems (Lennings & Pritchard 1998). Furthermore, factor analyses of the DSM-IV measurements also clearly indicate that the criteria do not discriminate dependency from frequency of use (Fulkerson et al. 1999).
A recent study of 4644 American arrestees and detainees aged 9 to 18 did not use any scale of dependency, but simply asked the youths if they had ever felt dependent on any substance (Yun Soo Kim & Fendrich 2002). Less than 10 per cent of the participants provided a positive response. This finding is difficult to reconcile with studies based on DSM criteria (McCelland et al. 2004; Lennings & Pritchard 1998). In fact, Lennings and Pritchard (1998: 151) themselves suggest that juvenile detainees, though they may abuse substances by using them frequently, are less likely to 'reflect' actual dependency because of their age. That is, by virtue of being young the detainees have not had the opportunity to develop substance dependencies in the same way as adults.
The responses to the Hoffman et al. (2003) scale were compared with the rates of substance use self-reported by the detainees (Tables A.4 and A.5, below). The responses to the Hoffman et al. (2003) scale were split into three groups. Youths who scored 0-2 were classified as low risk of dependency, 3-4 as moderate, and 5-6 as high.
The Hoffman et al. (2003) scale correlated very highly with frequency of use for both alcohol and drugs (p<0.01).
In summary, it was decided not to use the dependency scale in analyses for this report and instead present frequency of use. The results of the scale indicated very high levels of alcohol and drug dependence among the detainees. However, the dependency scale was not generated with a juvenile sample. Currently, the literature on adolescent dependency is confused and does not provide a benchmark with which to compare these results. In factor analysis, it seemed the scale measured one consistent construct. However, it is not clear whether this factor is dependency. Indeed, the high correlations with frequency of use may indicate that, for juveniles, the construct that the scale measures is actually frequency of use.
- Next section: References
- Previous section: Conclusions
- Contents