Scottish Government

Appendix 4: Representativeness of the Outcome Samples

The fact that fewer than half of those who were contacted actually took part in the study raises the question of how representative the interviewees are of the clients who are involved with the projects. This issue is particularly pertinent to the three young people's projects where the response rates were lowest. Unfortunately we can only cast a little light on this. Since we were unable to collect information on the characteristics of the potential respondents, it is not possible to make direct comparisons between participants and non-participants. We can, however, compare the gender and age profiles of the respondents from the East Ayrshire and Reiver projects with the age and gender profiles of all the clients with whom these projects had been in contact as of July (Reiver) and September (East Ayrshire) 2003 as detailed in the earlier profiles report (McIntosh et al 2004). Even so, it should be borne in mind that any difference between the profile of our respondents and that of all clients up to mid-2003 could reflect changes between then and spring 2004 in the characteristics of the people with whom the projects were in contact rather than differences between participants and non-participants in the study. The results of the comparison are shown in Table 3.

It appears that our sample of young people contains more female clients and fewer aged 12-14 than we might have anticipated. It therefore seems likely that male clients and younger teenagers were less likely to participate in the study. This might be a reflection either of the clients' own views or those of their parents whose non-response to the initial request for interview was the main reason for non-participation in the study. The effect that these biases might have had on clients' responsiveness to the projects is impossible to ascertain.

The problem of representativeness is less acute for the Aberlour sample. While 14 of the 15 respondents (93%) were female, this is consistent with the gender profile of this project's clients. As documented in our earlier report, 95% of the 139 clients who participated in the project between November 2001 and September 2003 were female (McIntosh et al, 2004). In addition, since the response rate for the Aberlour clients was 75%, there is considerably less reason to suppose that those who participated in the study are substantially different from those who are currently in contact with the project.