Despite global outrage at several widelypublicized,extreme incidents of sorcery accusation related violence (SARV) in Papua New Guinea (PNG), research into SARV has been largely limited to ethnographic accounts, with little done to document its prevalence or the responses that prevent or limit its occurrence. This paper describes an innovative and collaborative approach adopted to generate and integrate data for a mixed methods study of SARV. This project has built two significant new datasets and collected extensive qualitative data through interviews, focus groups, and participant observation in workshops and meetings. We describe our participatory, collaborative,and ethical approach, and why a mixed methods research (MMR) design was essential. The key data generated by the project is explained, with special attention given to the most innovative and vital element of the project: the complex and detailed incident data collection in selected locations. The subsequent section summarizes how the principles of grounded theory are helping to develop and revise conceptual and thematic strands across multiple sources of data, and the practical use of spatial-temporal coding to link and compare different sources of data. Several examples of preliminary findings are provided in order to illustrate the analytical advantages of the project's MMR design and collaborative approach. The final section acknowledges the limitations of the study design and the ways these are being mitigated.