Page 144 - PhD GT

6.5.5 Construct validity
For the purpose of examining construct validity, no scales were available specifically to measure coping strategies relevant to users of heroin. As mentioned in the validation of the Impaired Control Scale, factor analysis may be used to ascertain whether scale items measure the same construct. As discussed earlier, while 32 items had a factor loading greater than .4 on a main factor, this factor accounted for only 26.4% of the variance and the sample size was in any event too small for any conclusions to be drawn. The scale does, however, have good face validity and items clearly address methods of avoiding substance use.
6.5.6 Concurrent validity
For the purpose of examining concurrent validity of the scale, total scores and cognitive and behavioural sub-scale scores were compared with frequency and amount of heroin used, on the basis of the assumption that the more coping strategies were used to avoid consumption, the lower consumption would be. Correlations among measures of use, the total score for the coping behaviours scale, the cognitive sub-scale and the behavioural sub-scale were in the expected direction (ie negative) but did not reach significance. Correlations between coping and degree of dependence were also negative but did not reach significance.
6.5.7 Test re-test reliability
The adapted Coping Behaviours Inventory was subjected to a test re-test method for examining its reliability, while attempting to avoid the problems of the effects of resistance or of practice discussed by Oppenheim (1992) and addressed in the previous chapter. The separate sample used for the purpose of test re-test reliability was described as Sample 5 in Chapter 4. The adapted Coping Behaviours Inventory (CBIdrg) was administered in a battery containing two other questionnaires (being similarly subjected to tests of reliability in their own right) to 61 individuals of whom 57 completed this scale on two occasions, the first described below as t1 and the second as t2. Mean time between first and second scale completion, as already described in Chapter 5, was 5.3 days with a minimum of 2 days and a maximum of 14 days. Correlations and differences between the mean total scores for the two sets of responses are presented in Table 6.1.
138