Page 114 - PhD GT

high scores on Part 1 items while the latter group would obtain low scores on these items. Further, the authors predicted that those who had given up attempts at control would be distinguished from those who had never felt the need for control by the differences in their responses to items in Part 3 of the scale. Scale items in Parts 2 and 3 are identical, except that Part 2 refers to what has happened in the past six months and Part 3 items refer to beliefs about what would happen now. Items in these parts refer to the inability or difficulty in stopping once started, inability or difficulty in maintaining abstinence, drinking despite recognition of negative consequences, speed of drinking (as a behavioural correlate of impaired control), using external cues to terminate drinking and overall attempts to control drinking. The questionnaire was specifically designed to address the problems raised by Chick (1980b) regarding the presence or absence of intention to control which were discussed in an earlier section of this chapter.
When tested in an Australian clinical sample of 98 volunteers attending treatment services for alcohol related problems, the ICS was shown to have high levels of internal consistency and test-retest reliability (Heather et al. 1993). The results of principal components analysis indicated that each part of the scale was dominated by a single major component, that Part 1 (ICSAC) measured a different variable to Part 2 (ICSFC) and Part 3 (ICSPC), that Parts 2 and 3 measured similar but distinct variables. In order to examine the relationship between impaired control and the alcohol dependence syndrome, scores for Part 2 of the ICS and dependence scores as measured by the Severity of Alcohol Dependence Data (SADD) (Raistrick et al.1983) in the validation study (Heather et al. 1993) were subjected to principal components analysis. Results from the rotated solution showed two components which were separate but related. ICS scores loaded on a different factor to the majority of dependence scores with a moderate correlation between the sets of scores, leading the authors to conclude that a strong association exists between impaired control and the dependence syndrome:
“...that impaired control is related to a general factor of alcohol dependence but empirically distinct from it.” (Heather et al. 1993 p. 707)
Impaired control and inability to abstain did not form separate constructs, a finding similar to that of Kahler et al. (1995) who used multiple measures of dependence which contained both impaired control items and inability to abstain items.
The psychometric properties of the Impaired Control Scale were broadly confirmed in an English clinical sample (Heather et al. 1998) where Impaired Control scores were also shown to be
102