Zhe Cao, Lin Zhang, Zelin Wang, Li‐Qun Chen, Gunnar Sivertsen
The impact of scientific research beyond academia is receiving increasing attention from scientists, science policy, and society in general. However, the mechanisms driving this impact remain unclear. This study systematically reviews existing theoretical models and seeks empirical evidence from the REF impact case data. Through grounded coding analysis of 200 cases across four units of assessment, six stages of research generating impact were identified – problem identification, research conduct, output production, output utilization, impact formation, and multi-stakeholder interaction. These stages encompass 10 key elements and 43 corresponding sub-categories, forming a novel theoretical model. Comparative analysis revealed that, while the generation of research impact beyond academia generally follows a similar logic across disciplines, obvious differences exist in motivation, research approach, forms of outputs, types of impacts, etc. We provide four typical cases for illustration. Our findings contribute to a deeper understanding of impact generation processes for the societal benefits of scientific research, and provide a basis for research managers to assess scientific contributions according to their distinct characteristics.