A randomised controlled trial of a ‘checklist’ intervention to mitigate confirmation bias and improve forecasting accuracy in social work

Meindl, Melissa ORCID: https://orcid.org/0000-0002-1231-0175 and Wilkins, David ORCID: https://orcid.org/0000-0003-2780-0385 2026. A randomised controlled trial of a ‘checklist’ intervention to mitigate confirmation bias and improve forecasting accuracy in social work. The British Journal of Social Work , bcag212. 10.1093/bjsw/bcag212

PDF - Published Version Available under License Creative Commons Attribution. Download (1MB)

Abstract

Social workers may be particularly at risk of cognitive biases because they make judgements and decisions under uncertain conditions, involving predicting future outcomes. Given the potential consequences, evidence is needed on whether bias mitigation tools improve forecasting accuracy in social work. This study tested the impact of a checklist intervention on social workers’ forecasting accuracy (Brier scores) and confirmation bias (Wason Selection Task scores) using vignettes from real referrals. In an online survey, 88 participants responded to baseline vignettes before being randomly allocated to a control or intervention group. The intervention group completed the checklist before responding to endline vignettes and the Wason Selection Task, while the control group accessed the checklist afterwards. The intervention group’s mean Brier score decreased (0.44 to 0.41), indicating improved forecasting accuracy, while the control group’s increased (0.44 to 0.45), indicating reduced forecasting accuracy. These differences were not statistically significant. Wason Selection Task scores were similar across groups, indicating that the checklist made no difference to confirmation bias. Despite its effectiveness in other fields, a brief online checklist may not be sufficient to improve forecasting accuracy or confirmation bias in social work. Future research should explore practice-embedded or supervisory approaches in real practice settings rather than online.

Item Type:	Article
Date Type:	Published Online
Status:	In Press
Schools:	Schools > Social Sciences (Includes Criminology and Education) Schools > School of Social and Spatial Sciences Research Institutes & Centres > Children’s Social Care Research and Development Centre (CASCADE)
Publisher:	Oxford University Press
ISSN:	0045-3102
Date of First Compliant Deposit:	12 October 2026
Date of Acceptance:	9 February 2026
Last Modified:	12 Oct 2026 10:00
URI:	https://orca.cardiff.ac.uk/id/eprint/190128

Actions (repository staff only)

Edit Item