Page 143 - PhD GT

states that a minimum value of about .30 to .35 is required to consider that an item loads on any factor). The six items whose loading on this factor was less than .35 were all behavioural avoidance items (items 2, 10, 12, 18, 33 and 37); the highest factor loadings for these items were distributed amongst the remainder of the factors. It was difficult to interpret all the other factors, though one was made up of alternative positive activities and another behavioural avoidance items. The remaining factors accounted for between 6.5% and 2.7% of the variance. Since the sample size was unsuitable for factor analysis, having a subject : item ratio much smaller than 10 : 1 recommended by Nunnally (1978 p. 421), factor analysis was not pursued beyond examination of the question of whether there were two distinct factors accounting for cognitive and behavioural coping strategies respectively.
For the rest of the validation, the inventory was used precisely as that, an inventory of cognitive and behavioural coping strategies, consisting of both positive thoughts about giving up, negative thoughts about use, behavioural avoidance strategies and alternative activities inconsistent with using drugs. When the data were re-analysed into dichotomous categories to examine whether strategies had been used or not (as opposed to the frequency of their use), and divided into cognitive and behavioural coping inventories (see Appendix 13), the mean score for cognitive coping was significantly greater than the mean score for behavioural coping with a ratio of 4:3. The ratio of cognitive to behavioural items was computed from the frequency scores and the mean ratio was found to be 1.75:1. Only 5% of respondents used more behavioural than cognitive coping strategies. A significant positive correlation was also found between the use of cognitive and behavioural coping strategies.
6.5.4 Content validity
Repeated interviews conducted with participants at various stages of their addiction careers formed part of the process of examining the suitability of the CBI for adaptation to users of substances other than alcohol, as reported in an earlier section of this chapter. The scale was thought to be comprehensive and meaningful to respondents with drug and alcohol misuse disorders. As a result of these interviews, only two items were added to the scale due to the fact that they were not covered by existing items. In the event, one of these items referring to taking exercise, was used very little (mean item score .8 placing it among the five least used items) and the other item, referring to the taking of sensitising or substitute medication, is readily measurable by other means. It was therefore felt to be justified to delete these items for the purpose of comparability with the original CBI that was administered to the alcohol sub-sample in the main study.
137