Theory in academic research: field differences in uses

Purpose To explore new methods to gain some insights into how theory use varies between fields. Design/methodology/approach A triangulation of three muti-field partial studies of theory use. All studies use the full text of MDPI articles, a convenient science-wide source of clean article full texts, with the first analysing all large MDPI journals and the other two focusing on a multidisciplinary set of six. Study 1 used simple text matching of theory terms, and the others used large language models to extract theory-related information, also forcing them to name a “main theory” in each article. Findings Study 1 found large field differences in the extent to which theory is mentioned in MDPI journal articles, from “theory” rarely appearing in Biology and Life Sciences Forum (3%) to being common in Administrative Sciences (72%). The term “model” is more frequent, potentially signalling a theory-informed approach or an attempt at generalisation. Study 2 found substantial field differences in the explicitness, centrality and effectiveness of the main theory used in articles, according to Gemma3, with stronger contributions tending to occur in higher quality articles. Study 3 found substantial field differences in the commonness of individual named theories, the explicitness with which theory is referenced and the role that the main theory plays in articles. Practical implications Researchers should be sensitive to field differences in theory use, in case expectations learned from their home field are inappropriate elsewhere. Originality/value Although based on a single publisher, this study suggests that explicit invocation of theory is not the norm in most academic fields but that it is more typically implicit and unnamed – part of the fabric of a field.