There is a widespread consensus about the importance of social work supervision. However, we do not have a good sense of what works in supervision to make a difference for children and families.
This project involved training a small group of supervisors in a different approach to supervision (outcomes-focused supervision) and then measuring what difference it made for the supervisors, social workers and parents, compared to another group of supervisors who did not receive the training.
As a pilot feasibility study, we were not aiming to show a ‘treatment effect’. Instead, we wanted to explore the potential for testing this approach to supervision on a larger scale and to identify and test a range of outcome measures. We were successful in delivering the training and engaging supervisors and social workers, but less successful at involving families.
The aim of this project is to explore whether and how giving training to supervisors might work to make a positive difference for families, and to test the kinds of research design and outcome measures that would be useful in a larger-scale trial.
How we went about it
Working with Birmingham Children’s Trust, we identified two long-term child protection teams within the city. In one team, we trained supervisors to use outcomes-focused supervision. We did this via a combination of workshops and monthly action learning sets. In the other team, supervisors continued to provide supervision-as-normal. We audio-recorded supervision case discussions, interviewed social workers and interviewed parents to see what difference the training had made.
We blind-coded all the audio-recordings, to identify which case discussions showed more or less fidelity to an outcomes-focused approached.
This pilot feasibility study shows the potential for training supervisors, and how with a combination of workshops and monthly action learning sets, it is possible to support supervisors to behave differently in case discussions. We found that supervisors in the intervention group asked significantly more outcomes-focused questions than supervisors in the comparison group.
We also identified two outcome measures that looked most promising for a future large-scale trial. From interviews with parents, we found that those in the intervention group were more positive about their social worker in relation to agreement on tasks, agreement on goals and development of a good bond. We also found that on average, supervision case discussions in the intervention team were shorter and therefore less costly than in the comparison group.
In both groups, social workers were positive about the supervision they received, rating their supervisory ‘working alliance’ as very good and noting in interviews how accessible and supportive their supervisors were.
Future research would be needed to more robustly test these early findings. The main aim of this study was to pilot the training, monthly action learning sets and data collection techniques.
This study therefore shows the potential for testing this and other supervision training approaches using quasi-experimental or randomised controlled trial designs. It also indicates the promise of using outcome measures with parents that focus on measuring task agreement, goal agreement and bond, while also suggesting that to involve more families, we need to use more streamlined methods.
We will consider the findings and how they relate to the research priorities of the sector.