Page 146 - PhD GT

the present study, one main factor emerged in the scale and items loading on this factor were a combination of cognitive and behavioural coping strategies. Additional factors consisted of alternative positive activities items, behavioural avoidance items, positive cognitions about not using but these were not pure factors in the sense that not all items that could be defined in the same way loaded on the same factors and some of the factors were difficult to interpret at all. Given the insufficient size of the sample for the purpose of this analysis, coupled with the results that were found, it was not possible to say that there were two factors accounted for by cognitive and behavioural coping strategies which measured the same phenomenon. It was therefore decided, for the purposes of this study, to use the scale as two inventories, one for cognitive coping and one for behavioural coping.
Due to the length of the questionnaire, the question of item reduction was addressed. Spector (1992) suggests that one method of item reduction is to proceed on the basis of removing those items with item-total correlations less than .40. On this basis, ten items were removed: these were item numbers 2, 8, 10, 12, 14, 21, 23, 33, 35 and 37, all behavioural items with the exception of item 23. The alpha coefficient was raised by .003 as a result (to slightly over .92 as opposed to slightly under .92). As the difference rendered to alpha by item reduction was so small, but the removal of these items shifted the balance of items markedly towards a far greater proportion of cognitive items, the original set of 36 items (minus the two additional items) was used in the main analysis.
In the addiction outcome literature, outcome expectancies, self-efficacy, coping strategies and the perceived effectiveness of coping strategies all have predictive ability. It would appear, from examining the items generated in this small study, that the cognitive coping strategies identified were cognitions based upon a combination of negative outcome expectancies for continuing use and positive outcome expectancies for change. These two categories of coping strategies have been examined for their ability to predict outcome in people in treatment and not in treatment and have been found to be good predictors of outcome (Brown 1985, Jones and McMahon 1994). Jones and McMahon (1994) have argued that the research has focussed on positive outcome expectancies for drinking and their ability to predict relapse to the exclusion of attention to negative outcome expectancies and their ability to predict survival (or abstinence). Their study of negative outcome expectancies for drinking shows an equal ability to predict outcome in treatment and community samples (McMahon & Jones 1996). Though not all cognitive coping strategies can be clearly defined as positive and negative outcome expectancies in the current study, those cognitive items most frequently used are an equal number of positive thoughts
140