Page 158 - PhD GT

all correlations were significant at the .001 level). In order to examine its factor structure the scale was subjected to exploratory Principal Components Analysis which yielded two main factors accounting for 38.7% and 15.4% of the variance respectively. Further analysis with Varimax Rotation showed that the two factors comprised four items each. Items 1, 2, 7 and 8 (see Appendix 15) had loadings .55 and over on the first factor and items 3, 4, 5 and 6 had loadings of .59 and over on the second factor. The first factor appears to consist of aspects of home life while the second factor appears to consist of social aspects external to home life (employment, finance and social activities).
Concurrent validity was examined by correlating the two separate factors with dependence and psychopathology scores. The expectation would be that there would be co-variance but not large correlations given the hypothesised distinctiveness of the domains of dependence, psychological health and social satisfaction. This was found to be the case (Raistrick et al. in preparation). Increasing levels of social dissatisfaction as measured by both factors were associated with increasing levels of dependence and psychopathology. However, there was a negative correlation between age and factor two: the younger respondents expressed greater levels of dissatisfaction on items referring to employment, finance and social activities than did older clients. These findings were confirmed in a stepwise multiple regression analysis. Independent variables were age, gender, GHQ score, LDQ score and substance group and separate analyses were carried out for the two factors. Gender and substance group were not found to be independent predictors of social dissatisfaction, dependence and psychopathology scores were independent predictors of both factors and age was an independent predictor of the second factor.
A further test of the reliability of the scale was conducted with a separate sample using the test re-test method. This is the only part of the validation of the SSQ which I conducted as part of the present set of studies. The scale was administered as part of a package of three questionnaires to 61 respondents attending the addiction agency in Leeds for treatment of their problem heroin use (Sample 5 described in Chapter 4). The time frame for the collection of these data post-dated that for the large sample described above. It was a narrower sample in that it did not include problem drinkers. The re-test sample consisted of 40 males (65.6%) and 21 females (34.4%), their mean age was 25.7 (SD = 4.9) and a range of 17 to 37. The mean time between first and second scale completion was 5.3 days (SD = 3.75) with a minimum of two days and a maximum of 14 days.
The study in which these data were collected was Study 3 (adaptation and validation), the retest reliability part of the study reported in Chapters 5 and 6.
A significant correlation was found between the mean SSQ scores derived at the two data 152