2 Research methods
This research used a mixed methods approach, comprising questionnaires, documentary analysis, face-to-face interviews and focus groups with key stakeholders in the children’s hearings system, including sheriffs, panel members, social workers, solicitors, reporters, managers from Children 1 st (the organisation contracted by the Scottish Government to manage the national Safeguarders Panel) and safeguarders.
This Chapter details ethical issues as well as the access arrangements to, and demographics of, the various samples drawn on in the study, including the ways in which the samples were selected and more detail on the way in which methods were applied.
This work was conducted in line with the University of Strathclyde Ethics Committee’s Code of Practice following ethical approval by the Law School Ethics Committee UEC16/71.
Electronic data: all electronic data were stored on secure servers based at the University of Strathclyde, Glasgow. The data files were accessed using password – protected computers by members of the research team. User specific permissions were used to limit data file access to the appropriate member of the research team.
Interview and focus group data: interview recordings were transferred from the recording device to a password-protected computer, after which the original recording was permanently deleted. A professional transcriber, subject to a confidentiality agreement, was used to transcribe the interviews prior to analysis.
Sample and paired report data: anonymised records of cases where safeguarders have been appointed by hearings and sheriffs was obtained from both SCRA and Children 1 st. Data related to the paired social work and safeguarder reports were made available by a local authority following separate ethical approval. Data were extracted manually by a member of the research team and anonymised at the time of extraction.
Informed consent was received at each stage of the project and all data will be permanently deleted within five years.
2.3 Project phases
2.3.1 Phase 1: Scoping interviews
To help inform the latter phases of the project and arrange/address access, scoping interviews were conducted with a sheriff and five senior managers, one from each of the key stakeholder groups: Children 1 st, Children’s Hearings Scotland, Scottish Children’s Reporter Administration ( SCRA), the Scottish Legal Aid Board ( SLAB) and a local authority social worker.
2.3.2 Phase 2: Questionnaires
The online survey software Qualtrics was used to gather opinions on the role of the safeguarder from 472 individuals from various stakeholder groups across Scotland ( Appendix 1). The various agencies involved in overseeing or administering the work of the relevant stakeholders i.e. SCRA (children’s reporters), Scottish Courts and Tribunals Service ( SCTS) (sheriffs), Children 1 st (safeguarders), SLAB (solicitors), Children’s Hearings Scotland (panel members) and Social Work Scotland (social workers) assisted the research team in identifying ways of advising the stakeholders that the questionnaire was available. A modified version of the questionnaire was used to collect responses from sheriffs ( Appendix 1). The questionnaires used both open and closed question types, such as: scales (i.e., extent of agreement on a scale of 0 – 10), categories (i.e., yes/no), and free-text responses, the latter of which were coded to enable more in-depth analysis.
Summary information on questionnaire respondents
Ninety-nine safeguarders (21%), 357 non-safeguarders (77%) and 16 sheriffs (2%) responded.
In terms of the demographics for safeguarders, the most common categories were: female gender (n = 64, 65%), age 60 – 69 years (n = 48, 48%), never been a panel member (n = 86, 87%), had more than 10 years’ experience as a safeguarder (n = 39, 39%), fulfilled the role in 3 areas (n = 30, 30%) and retired as main occupation (n = 42, 44%) ( Appendix 2 Tables 201 – 207).
For non-safeguarders, the most common categories were female gender (n = 250, 70%), age 50 – 59 years (n = 95, 27%), main role in children’s hearings process was panel member (n = 145, 47%) and fulfilled the role in one area (n = 278, 78%) ( Appendix 2 Tables 208 – 215). Nine of the non-safeguarders had previously been safeguarders.
For sheriffs, the most common categories were 10 years working as a sheriff (4/13), had appointed safeguarders (12/13) and had safeguarder involvement in 1 – 10 cases before them (5/13) ( Appendix 2 Tables 213 – 215).
All responses were analysed using the statistical software package Minitab Version 17.
2.3.3 Phase 3: Documentary analysis
This phase involved the analysis of cases where safeguarders had been appointed within the past 24 months.
A sample of 50 cases in which safeguarders had been appointed by sheriffs (hereinafter “the sheriff sample”) and 50 cases in which safeguarders had been appointed by children’s hearings (hereinafter “the SCRA sample”) were selected to examine the reasons for safeguarder appointments. Whilst the sheriff sample was restricted to reasons for safeguarder appointments, much more information was provided in the SCRA sample. This allowed the research team to “track” the 50 cases in the SCRA sample through the children’s hearings process from safeguarder appointment to substantive decision and, where appropriate, appeal.
A separate sample of safeguarder reports was also analysed alongside, and compared with, the corresponding social work reports in 17 individual cases (hereinafter referred to as the "paired report analysis”).
Selecting reasons for appointment records
The sheriff and SCRA samples were taken from three areas in Scotland. These areas were chosen by reference to the Children 1 st Safeguarder Panel Annual Report 2015/16  because of the contrasting levels of safeguarder appointments made by sheriffs versus hearings: one area had a high sheriff appointment rate, one had a high hearings appointment rate, and one had a relatively equal number of both sheriff and hearings appointments. The paired report sample was drawn from one of these areas.
The SCRA Statistical Analysis 2015-16  was used to ensure that the sheriff and SCRA samples had an age and gender distribution consistent with these overall Scottish statistics. An overall set of records was identified in both cases by application of these criteria (ie area, gender and age of child, 24-month timescale) and the appropriate number of records with the specific characteristics was randomly sampled. The main aim of this exercise was to provide a comprehensive, albeit not representative, overview of the specific reasons for which safeguarders were appointed by children’s hearings and sheriffs.
The sheriff sample
For the sheriff sample, the SCTS indicated that sheriffs’ reasons for safeguarder appointments were unlikely to be specified in court documents, and a request was therefore made to Children 1 st for access to anonymised safeguarder allocation forms. Fifty such forms were identified as above. The variables identified in Appendix 2, Table 216 were extracted for analysis (although, almost all of these forms merely supplied the rationale for sheriff appointments of safeguarders).
Twenty-four (48%) of the sheriff sample were male and 26 (52%) were female. The ages ranged from 2 weeks to 15 years, with 20 (40%) in the range 0-5 years; 12 (24%) 6-10 years and 18 (36%) 11-15 years. In 28 (56%) of the sheriff sample, more than one child was included in the referral to the safeguarder. There was no indication as to the grounds of referral but 47 (94%) related to proof proceedings, with the remainder relating to appeals.
The SCRA sample
The SCRA sample was collected and anonymised by SCRA. These data not only included reasons for appointment, but also demographic information on the child(ren) to which cases applied, grounds of referral, type of order and measures in place, purpose of hearing at which a safeguarder had been appointed, the safeguarder’s recommendation, the substantive decision and any appeal outcomes. Significantly more data was collected in this sample than the sheriff sample. A much richer analysis was therefore possible for the SCRA sample and the data sets are not comparable, other than information collected pertaining to the stated reason for the appointment of safeguarders by hearings and sheriffs. Data was aggregated under four headings or stages: background; safeguarder appointment; substantive hearing decision; and, appeals. The variables identified in Appendix 2, Table 217 were extracted for analysis.
Twenty-four (48%) of the sample were male and 26 (52%) were female. The ages ranged from 1 year to 16 years, with 18 (36%) aged 0-5 years, 16 (32%) aged 6-10 years and 16 (32%) aged 11-16 years. The vast majority (n = 47, 94%) were already subject to a formal order, most often a Compulsory Supervision Order ( CSO) (n = 34, 68%). Ten (20%) were subject to interim measures and a few were subject to a Child Protection Order. Children subject to compulsory measures prior to safeguarder appointment had been on these from between two days and nine years.
A single accepted/established ground of referral was listed in 36 (72%) of the cases, 2 such grounds applied in 11 (22%) cases and 3 in 3 (6%) cases. Overall 38 (76%) cases related to the lack of parental care ground. All accepted/established (as opposed to new) grounds were care and protection rather than related to offending.
Paired report analysis
A sample of 17 anonymised paired reports, one of each pair having been submitted by the safeguarder and the other by the social worker for the same child in the same case were selected in order to compare their style, structure, content and recommendations as well as provide a better understanding of any added value provided by the safeguarder reports. Analysis was done manually, using a coding sheet and the variables identified in Appendix 2, Table 218 were extracted for analysis.
Nine (53%) of the sample were male and 8 (47%) were female. The ages ranged from 1 year to 13 years. The majority (n = 11, 65%) were aged 6-10 years. Twelve (71%) related to a single child, rather than multiple siblings. Sixteen (94%) of the 17 reports related to children who were already subject to CSOs and all related to care and protection grounds rather than offence grounds. The majority (n = 15, 88%) were ‘looked after and accommodated’ by the local authority.
2.3.4 Phase 4: Interviews
In November and December 2016, 38 interviews with key stakeholders were conducted across Scotland to explore key topics around the safeguarder role, and individual professionals’ views and experiences of safeguarders. Interviewees were selected from a group of 138 questionnaire respondents who consented to participate in this phase.
Those people who had completed a questionnaire were asked if they would consider being approached for a one to one interview or to participate in a focus group discussion at a later stage in the fieldwork process. This generated a list of 40 safeguarders, 5 solicitors, 11 reporters, 57 panel members and 25 social workers. From this, the research team identified safeguarders and panel members for interview by using a random selection process based on the Allocation Table contained in the Children 1 st Annual Report 2015/16 as a guide to the national picture. We then identified a random mix of potential interviewees across the highest, median and lowest allocating local authorities, based on gender and length of service. Social workers and reporters were broadly identified in the same manner. With sheriffs, we were confined to two sheriffdoms, one with a high and one with a median allocation of safeguarders and Sheriffs Principal facilitated identification of interviewees.
We interviewed 11 safeguarders (including one pilot interview); 10 panel members; 5 social workers; 5 reporters; 5 solicitors; and 2 Children 1 st staff. Interviews with 9 sheriffs took place between January and March 2017. Interviews mainly took place in people’s offices and lasted between 60 and 90 minutes; average length of interview was between an hour and an hour and a half.
2.3.5 Phase 5: Focus groups
Three focus groups – one each with safeguarders, panel members and social workers – were undertaken in April 2017. The safeguarder focus group, held in Edinburgh, included 3 males and 2 females, working in more than 6 areas. The panel member focus group, also held in Edinburgh, included 6 males and 3 females, working in more than 7 areas. The social worker focus group, held in Glasgow, included 4 males and 5 females, some of whom were team leaders, and all of whom worked in one area.