Scientific methodologies for interdisciplinary research: an empirical solution to the demarcation problem

Background Scientific interdisciplinary research must involve disciplines which share a factually common core of scientific methodological features. Methods This paper suggests a theoretical framework to solve the demarcation problem, by referring to inductive vs. deductive/abductive, topical vs. contextual and experimental vs. observational properties within the cognitive-epistemic aspects of science. It compares alternative H-indices (from H-index based on articles to net per-capita per-year H-index based on articles) as common outcomes of scientific activity to measure different scientific methodologies characterising different disciplines and subjects, by constructing a representative dataset based on a cross-disciplinary balanced sample (10,000 authors with at least one article indexed in 2006-2015 Scopus covering 25 disciplines and 4 subjects). It validates this distance metric for both disciplines and subjects, by using it to highlight the overall methodological difference between disciplines and subjects. Results Subjects are statistically comparable, apart from a significant difference between Physical vs. Social Sciences (they differ for all six methodological features) and a slightly significant difference between Life vs. Health Sciences (topical vs. contextual) and Life vs. Social Sciences (topical vs. contextual and experimental vs. observational). Most disciplines are statistically comparable, apart from a significant difference between Physics & Astronomy and Chemistry vs. some other disciplines (the significant dissimilarity observed for subjects is due to the peculiarity characterising these two disciplines). Conclusions A continuous range of scientific methodologies is empirically identified, by characterising interdisciplinary science as referring to six properties to different extents and spanning from Arts & Humanities to Physics & Astronomy. The validated theoretical framework enables the separation of scientific from non-scientific articles (although each article can be characterised by different degrees of methodological properties) as well as the separation of scientific from non-scientific publication history (although each publication history can be characterised by different percentages of methodological features).