Despite the widespread adoption of Theories of Change (ToC) for programme evaluation, the process of collaboratively developing these theories is rarely outlined or critical analysed, limiting broader methodological discussions on co-production. We developed a ToC as part of E le Sauā le Alofa ('Love Shouldn't Hurt') - a participatory peer-research study to prevent violence against women (VAW) in Samoa. The ToC was developed in four phases: (1) semi-structured interviews with village representatives (n = 20); (2) peer-led semi-structured interviews with community members (n = 60), (3) community conversations with 10 villages (n = 217) to discuss causal mechanisms for preventing VAW, and (4) finalising the ToC pathways. Several challenges were identified, including conflicting understandings of VAW as a problem; the linearity of the ToC framework in contrast to intersecting realities of people's lived experiences; the importance of emotional engagements, and theory development as a contradictory and incomplete process. The process also raised opportunities including a deeper exploration of local meaning-making, iterative engagement with local mechanisms of violence prevention, and clear evidence of ownership by communities in developing a uniquely Samoan intervention to prevent VAW. This study highlights a clear need for ToCs to be complemented by indigenous frameworks and methodologies in post-colonial settings such as Samoa.
Keywords: Samoa; Violence against women; theory of change; theory-driven approaches to intervention development.