Page 234 - PhD GT

p. 301). In the first of these three methods, the cut-off point for asserting that a move from the dysfunctional to the functional has occurred is said to be reached with scores that fall outside the range of the dysfunctional behaviour, where the range is described as being two standard deviations in the direction of improvement.
To apply this method in order to meet the criterion “that clients end up within the range that renders them indistinguishable from well-functioning people”, the calculation of clinically significant change would be based upon the mean dependence scores for the study population or a similar clinic attending population. Mean dependence score for the entire study population (n = 230) was 20.2 with a standard deviation of 7 (see Chapter 8). When the sample was divided by substance, the heroin group had a mean dependence score of 21.2 with a standard deviation of 6.3 and the alcohol group had a mean dependence score of 18.2 with a standard deviation of 7.8. In a larger sample of people attending both the same agency at a different time and a similar agency in a different geographical population (n = 1681, Heather et al. submitted), the mean dependence score for the entire sample was found to be 19.7 with a standard deviation of 7.6; the heroin sub-sample had a mean dependence score of 21.6, standard deviation 6.7 and the alcohol sample had a mean dependence score of 18.4, standard deviation 8. The similarity in these sample means justified the use of either as being representative of dysfunctionality in this sort of clinic attending population. The cut-off point calculated for significant change (on the basis of scores for the study sample) would be a mean dependence score of 6.1 for the whole group or 8.5 for individuals with heroin dependence and 2.6 for the individuals with alcohol dependence.
In the second method, the cut-off point is reached when scores fall within the functional range where the range is set at two standard deviations above (in the case of the measurement of dependence) the measurement for the normal population. If the second method were to be used, a general population measure of dependence as measured by the LDQ was required. Such a measure was reported in the validation of the LDQ, albeit in a very small (n = 14) general practice sample, to be a mean total score of 3.1 with a standard deviation of 3.2 (Raistrick et al. 1994). If this rough measure were used as a bench mark, examination of change would be focussed on those participants whose t2 and or t3 scores for dependence were 9.5 and less. Thus, in line with the findings reported by Jacobson et al. (1999), the second method was found to be the less stringent of the two.
The third method requires that the individual is statistically significantly more likely to belong to the functioning group than the dysfunctional group, and the sample from which normative dependence data could be derived for this purpose, as described above may not be sufficiently large. Thus one could establish with greater certainty that outcome measures were statistically
229