Page 188 - PhD GT

p<.01) from the correlation between these total dependence scores for the alcohol sub-sample (correlation coefficient .7, see Table 9.1). Furthermore, the correlation between total dependence score at intake and twelve months in the heroin sub-sample (correlation coefficient -.08) was significantly different (z=1.788; p<.05) from the correlation between these total dependence scores in the alcohol sub-sample (correlation coefficient .29). The correlation between total dependence score at three months (LDQ t2) and at twelve months (LDQ t3) for the heroin sub-sample (correlation coefficient .03) was significantly different (z=2.327; p<.01) from that of the alcohol sub-sample (correlation coefficient .48).
In order to examine change in dependence, a change score was computed for the difference between dependence at intake and three months (t1-t2), the difference between dependence at three months and at twelve months (t2-t3) and the difference between intake and twelve months (t1-t3). Change in dependence scores between t1 and t2 was a mean reduction of the order of 7 points on the LDQ, the median change was a reduction of 5.5 points. Change in dependence scores between t2 and t3, in other words, between three and twelve months was a mean reduction of the order of 1.2 points and the median reduction score was 0.
When the mean of the change scores between t1 and t3 was computed, a mean reduction of 8.4 points was found with a median reduction score of 7 points. The mode was a reduction of 5 points. Mean change scores and their distributions are shown in Table 9.2. There were no significant differences between the drug groups in the amount of change in dependence between intake and three months, three months and twelve months or intake and twelve months.
Individual differences in the amount of change were considerable as suggested by the common range for each of the follow-up periods, compared to the mean change scores. Between t1 and t2 for example, 10% reduced their dependence score by only one point while 20% reduced their dependence score by 14 points or more (up to 27 points).
Also, the nature of change both within and between individuals appears quite different for the first three months compared with the following nine months. The proportions of respondents who reduced their mean dependence score, who increased their mean dependence score and who did not change their mean dependence score were examined in order to trace the course of change between t1 and t2, and t2 and t3; the findings are shown in Table 9.3.
183