
Episode 33 – Realist Evaluation in Commissioned Settings with Brad Astbury
What does it mean to conduct realist evaluation well when timelines are tight, budgets are limited, data access is imperfect, and commissioners want findings that areclear, useful, and actionable? How can evaluators remain faithful to realist principles without becoming trapped in theoretical complexity, or reducing context, mechanism, and outcome configurations to formulaic boxes? In this refreshingly practical conversation, Alejandro is joined by Associate Professor Brad Astbury, Associate Professor of Health Systems Evaluation at the University of Melbourne, whose work spans evaluation practice, public health,government-commissioned evaluation, programme theory, teaching, and realist approaches. Brad reflects on his journey from criminology and the corrections system to more than two decades of engagement with realist evaluation. Drawing on his extensive experience conducting evaluations for government, he explores some of the most common ways realist approaches are misunderstood, overcomplicated, underdeveloped, or applied in circumstances where another evaluation approachmight be more appropriate. The discussion explores the pressure commissioners and public servants face to produce straightforward answers to complex questions. Brad explains how evaluators canacknowledge legitimate accountability needs while also generating the more nuanced knowledge required to improve, adapt, target, transfer, and scale programmes. Throughout the episode, Brad emphasises the importance of accessible language, listening before prescribing a methodology, preparing to explain realist ideas todifferent audiences, and approaching evaluation as a form of disciplined curiosity that asks “why?” relentlessly. He concludes with four practical principles for commissioned work. Whether you are an evaluator, commissioner, researcher, public servant, policymaker, consultant, student, or PhD researcher, this episode offers practical wisdom for conducting realist evaluation with integrity under real-world conditions.