The project adopted a mixed methods approach involving both quantitative and qualitative methods which generated a particularly rich data set. Scoping interviews were initially conducted with one senior individual with particular responsibility in relation to each of the six stakeholder groups (eg senior manager at the Scottish Legal Aid Board in respect of solicitors). This helped to identify key issues in relation to the safeguarder role. An online questionnaire was then distributed to members of five stakeholder groups. In the end, 472 responses were received (from 99 safeguarders, 357 non-safeguarders (comprising reporters n = 41, 13%; lawyers n = 16, 5%; panel members n = 145, 47%; social workers n = 85, 28%; and others n = 20, 6%). A tailored version was separately distributed to sheriffs and 16 responded. All responses were analysed and coded to provide quantitative data on stakeholders' views.
The third phase involved documentary analysis of 50 sets of reasons for appointment given by children's hearings ("the SCRA sample") and 50 sets of sheriff's reasons ("the sheriff sample"). Whilst the sheriff sample was restricted to reasons for safeguarder appointments, much more information was provided in the SCRA sample allowing "tracking" of the 50 cases from safeguarder appointment to substantive decision and a richer overall analysis. Phase 3 also included a separate analysis of 17 pairs of social work and safeguarder reports in the same case ("the paired report sample"), allowing an insight into reasons for which safeguarders are, in practice, appointed and also some indication of convergence with, or divergence from, social work recommendations. The paired reports were examined on their own, rather than within the case papers and could not, therefore, be correlated with the substantive decision taken in these cases.
Phase 4 involved semi-structured interviews with 47 stakeholders (9 sheriffs, 10 panel members, 11 safeguarders, 5 reporters, 5 solicitors, 5 social workers and 2 Children 1 st managers) allowing a more in-depth examination of their views. The final phase consisted of three focus groups (one for safeguarders, one for panel members and one for social workers) at which preliminary findings were discussed. Interviews and focus group discussions were recorded and transcribed and then coded and analysed to identify key themes and to triangulate responses in other phases.
All data were anonymised and electronic data were held in accordance with data protection requirements. While the data yielded were rich and varied, all of the sample sizes were small and not representative, and the qualitative elements reflect the views only of the small numbers of stakeholders involved, so that findings should be approached with this in mind and interpreted with caution.