Developing a critical imagination for how researchers can use artificially intelligent tools reflexively and responsibly during qualitative literature reviews

Texto completo

Artificially intelligent (AI) tools hold significant potential for researchers during the processes of qualitative literature review. Beyond only literature retrieval, some AI tools describe their use-value as being able to appraise and synthesise contrasting evidence within and across papers; work like «Spotify» to cultivate a personalised ecosystem of research articles that evolves via algorithmic learning; and suggest new directions for research based on existing evidence about a topic and an assessment of knowledge gaps. Our manuscript engages with this critical challenge by using multi-layered forms of reasoning to develop a critical imagination for qualitative researchers integrating novel forms of AI into their literature review methodology. We outline the roles and functions of ten AI tools that support researchers to map, synthesise, and appraise knowledge — using the citation environment (Citation Tree, Semantic Scholar, Open Knowledge, Connected Papers, Inciteful), article content (Litmaps, VOSviewer), and those suggested capable of assessing the literature to summarise key terrain or suggest next steps for inquiry (SciteAI, Elicit, Research Rabbit). We discuss ethical implications of the uncritical use of AI tools during review of literature and delineate essential reflexive responsibilities.