Abstract

Despite widespread calls to make prisons more rehabilitative, issues such as aggression, substance use, and recidivism remain pervasive. Evidence-based practices in correctional settings have struggled to deliver consistent improvements, partly due to weak theoretical and empirical foundations. Many programs rely on theories focusing on distal outcomes like recidivism that do not specify how behaviour arises from interacting goals, beliefs, and contexts.. This review examines the stagnation of evidence-based rehabilitation and how to enhance it by grounding practice in contemporary psychology. We identify three challenges: the replication and generalisability crisis, the mapping problem between theory and outcomes, and structural and institutional barriers to implementation. Building on recent cognitive accounts of behaviour, we outline a framework that reconnects behavioural evidence with the processes it seeks to change. We conclude with recommendations for practice: supporting adaptive inference chains, strengthening coherent self-models, aligning environments with predictive learning, and embedding psychological expertise across institutional levels.

Keywords: evidence-based prison policy, rehabilitation, recidivism, psychology, behaviour change, goal-directed predictive processing

How to Cite:

Linthout, T., Caspar, E. & Van Dessel, P., (2026) “Learning to Change: A Goal-Directed Perspective on Evidence-Based Rehabilitation”, Law & Criminology Journal 3(1), 168-189. doi: https://doi.org/10.21825/lcj.99704