Liz Beddoe| Gillian Ruch| Synnöve Karvinen-Niinikoski |Ming Sum Tsui
Supervision is a topic of considerable interest in social work at present. Supervision is seen as a vital activity for meeting many professional demands: the continuing development of professional skills, the safeguarding of competent and ethical practice and the oversight of casework. A strong research base is necessary to ensure that there is empirical support for supervision as a core practice in social work (O’Donoghue & Tsui, 2013). A recent report reviewed the evidence for supervision: “Effective supervision in social work and social care” by Carpenter, Webb, Bostock and Comber (2012) which is available here A key message from this project was that “overall, the empirical basis for supervision in social work and social care in the UK is weak. Most of the evidence is correlational and derives from child welfare services in the US” (p.1). This is a significant challenge to those of us who are committed to teaching and research about supervision. The project team sought to access the views of colleagues with and interest in supervision research and development in social work and a Delphi study seemed to offer a fruitful method.
The Delphi Study is a much utilised study method for establishing a consensus (specialist / regional / international) consensus on subjects such as research priorities and best practice guidelines. The Delphi method is frequently used in nursing research. A Delhi study is essentially a multi-phase project involving two or more questionnaires to “experts” or important stakeholders in which each iteration produces clearly ranked research or other priorities. The first questionnaire presents the problem or focus for the discussion and collects suggestions from invited participants. The first round of data is collated and used to create a second questionnaire, which allows participants the opportunity “to re-evaluate their responses in light of those of others, then rank the items” (Wathen, MacGregor, Hammerton et al, 2012, p. 2). The aim of the second round questionnaire is to develop a broad consensus amongst a community of interest and disseminate a set of priorities. For this project, team members hope that one of the outcomes of the project will be some greater engagement of social work supervision researchers in multi-country research and development collaborations.
One issue that emerges in the literature reporting Delphi studies is that of choosing the “experts” to contribute to the research data. West (2010) suggests that in many studies these are solely drawn from academic institutions and that this may limit access to the views of a rather narrow elite group to achieve the Delphi consensus ideal. Indeed “quality” of the experts is frequently mentioned as a concern in assessing the rigour of Delphi study results (Powell, 2003). On the other hand the Delphi study can provide the opportunity to draw a broader consensus. Murphy et al. (1998) argued that diversity in the participant panel may lead to improved outcomes as it allows for input from different perspectives and experiences.
Our aim is to recruit to the study people with academic expertise in supervision (having published in the field or closely related material for example) and people whom we might define as expert users; for example those who are very involved in supervision as expert practitioners, practice teachers, trainers and those who might be influential in developing and implementing supervision policies within social service organisations. The advantage of this approach (along with the obvious one of inclusion of practice perspectives and those closer to direct practice) is that it enables us to contrast the views across these two groups. This adds an interesting dimension to the study.
Powell (2003, p.376) reviews the technique and concluded: “Although the technique should be used with caution, it appears to be an established method of harnessing the opinions of an often diverse group of experts on practice-related problems”. West (2010) and Wathen et al (2012) have reported good result for establishing research priorities and there are other similar studies.
A Delphi sample does not aim to be representative; rather it aims to bring knowledgeable and committed participants into a process of pooling ideas. There are debates about the desirable size of the participant pool (Powell, 2003). The team for this current project decided that it would be desirable to start with a reasonable large group –to assist with potential attrition at each iteration of the questionnaire. Our aim is to develop a pool of approximately 100-150 people. Countries for inclusion: while the pool of countries is ever growing at the time of writing this summary invitations will be sent to researchers from New Zealand, Australia, Ireland, England, Northern Ireland, Scotland, Wales, Hong Kong, China, Taiwan, Finland, Sweden, Denmark, Norway, the United States, Canada, South Africa, Israel, Botswana and South Korea.
Delphi Study Process
Phase One: Conduct a literature review based on publications on social work supervision over the past 6 years (2008-2013 inclusive) which has been reviewed for content relating to author’s recommendations for further supervision research . This review provides both information about the potential pool of participants and a set of data about published researchers’ thoughts on the need for further research.
Phase Two: Writing the study protocol, designing web survey gaining ethics approval from the University of Auckland. An online survey has been developed and a pool of potential participants has been established, based on the survey of the research literature and researcher professional networks. The survey includes some items designed to capture a profile of the participant panel with some information about a range of practices across different countries.
Phase Three: the research team will analyse the round one data and develop themes to explore further in the round two questionnaire which will ask participants to rank items related to research priorities for social work supervision.
Send out survey and collate and process data.
Phase Four: Testing the final consensus, which may involve and small invited advisory group. Dissemination of the results in publication and scholarly conferences.
Liz Beddoe PhD,
On behalf of the Delphi study team
Carpenter, J., Webb, C. M., Bostock, L., & Coomber, C. (2012). Effective supervision in social work and social care. Research Briefing 43. London: Social Care Institute for Excellence.
Murphy M.K., Black N., Lamping D.L., McKee C.M., Sanderson, C.F.B., Askham J. et al. (1998) Consensus development methods and their use in clinical guideline development. Health Technology Assessment 2(3), i-iv, 1-88.
O’Donoghue, K., & Tsui, M.-s. (2013). Social work supervision research (1970–2010): The way we were and the way ahead. British Journal of Social Work. 10.1093/bjsw/bct115
Powell, C. (2003). The Delphi technique: myths and realities. Journal of Advanced Nursing, 41(4), 376-382. doi:10.1046/j.1365-2648.2003.02537.x
Wathen, CN, MacGregor, JC, Hammerton, J, Coben JH, Herrman H, Stewart DE, & HL, M. (2012). Priorities for research in child maltreatment, intimate partner violence and resilience to violence exposures: Results of an international Delphi consensus. BMC Public Health 12:684 http://www.biomedcentral.com/1471-2458/12/684
West, A. (2010). Using the Delphi Technique: Experience from the world of counselling and psychotherapy. Counselling and Psychotherapy Research, 11(3), 237-242. doi:10.1080/14733145.2010.492429