Page 79 - PhD GT

significantly higher levels of heroin consumption than their New York counterparts, a ceiling effect on the consumption responses included in the SODQ might account for the different findings in the two countries.
Burgess et al. (1989) reported on further validation of the SODQ in an Australian sample. On this occasion the criterion measure used in the validation was a clinical rating of dependence based upon the Psychoactive Substance Dependence and Abuse section of the Structured Clinical Interview (SCID-R) (Spitzer et al. 1986) for the Diagnostic and Statistical Manual for Mental Disorders, Third Edition, Revised (DSM-III-R) (American Psychiatric Association, 1987). On this occasion the SODQ is described as consisting of eight sections (Burgess et al. 1989, p. 1451), having included the original ‘additional questions’ (Sutherland et al. 1986), namely those referring to demographic features and history of opiate use, narrowing of repertoire and tolerance in the main body of the scale. The findings in this study were consistent with previous validation studies leading the authors to question whether the rapid reinstatement element could be considered as part of the dependence syndrome or not. They further noted that some of the discrepancies found in examining correlation between SODQ scores and other measures of opiate dependence, notably those derived from the SCID-R (Spitzer et al. 1986) may be attributable to the way in which syndrome elements are operationalised in different scales. They dispute whether the items purporting to measure salience in the SCID-R are operationalised in line with their theoretical formulation, thus implying that they may be measuring something else. They apply the same criticism to items measuring re-addiction liability, although it would seem that such items have constituted a problem in other scales. Further criticism relates to the fact that neither scale covers all elements of the syndrome and there remains the question of whether therefore they can be said to test it and, as noted in an earlier context, they criticise their own study for using instruments for validation which are themselves not validated. These points raise concerns about the problem of attempting to define a phenomenon through its measurement and factor analysis.
3.4.6 Measuring substance dependence: the Severity of Dependence Scale (SDS)
In the following sections, two instruments designed to measure dependence on substances in general, the Severity of Dependence Scale (SDS) and the Leeds Dependence Questionnaire (LDQ), are described. The question of validation with reference to specific substances is addressed. Greater attention is paid to the LDQ as it is the main instrument used for measuring dependence in the present study. Finally some criteria for comparing scales and assessing their utility are proposed.
67