A workflow for collecting and understanding stories at scale, supported by artificial intelligence

Abstract

This article presents an artificial intelligence-assisted causal mapping pipeline for gathering and analysing stakeholder perspectives at scale. Evidence relevant to constructing a programme theory, as well as evidence for the causal influences flowing through it, are both collected at the same time, without the evaluator needing to possess a prior theory. The method uses an artificial intelligence interviewer to conduct interviews, automated coding to identify causal claims in the transcripts, and causal mapping to synthesise and visualise results. The authors tested this approach by interviewing participants about problems facing the United States. Results indicate that the method can efficiently collect and process qualitative data, producing useful causal maps that capture respondents’ views. The article discusses the potential of this approach for evaluation, enabling rapid, large-scale qualitative analysis. It also notes limitations and ethical concerns, emphasising the need for human oversight and verification.