The reliability of biomedical science: A case history of a maturing experimental field. (7th April 2022)

• Record Type: Journal Article
• Authors: Michell, Robert H.
• Is Part Of: BioEssays. Volume 44:Number 6(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-04-07
• Subjects: cancer -- cell signalling -- fatty liver -- fraud -- PCOS; Molecular biology -- Periodicals; Cytology -- Periodicals; Developmental biology -- Periodicals; 572.8
• DOI: 10.1002/bies.202200020
• Languages: English
• ISSNs: 0265-9247

Abstract

Abstract: There is much discussion in the media and some of the scientific literature of how many of the conclusions from scientific research should be doubted. These critiques often focus on studies – typically in non‐experimental spheres of biomedical and social sciences – that search large datasets for novel correlations, with a risk that inappropriate statistical evaluations might yield dubious conclusions. By contrast, results from experimental biological research can often be interpreted largely without statistical analysis. Typically: novel observation(s) are reported, and an explanatory hypothesis is offered; multiple labs undertake experiments to test the hypothesis; interpretation of the results may refute the hypothesis, support it or provoke its modification; the test/revise sequence is reiterated many times; and the field moves forward. I illustrate this experimental/non‐experimental dichotomy by examining the contrasting recent histories of: (a) our remarkable and growing understanding of how several inositol‐containing phospholipids contribute to the lives of eukaryote cells; and (b) the difficulty of achieving any agreed mechanistic understanding of why consuming dietary supplements of inositol is clinically beneficial in some metabolic diseases. Abstract : We have a growing understanding of exactly how inositol phospholipids contribute to several cellular functions, but we still do not understand why eating supplementary inositol ameliorates some metabolic diseases. This illustrates an important difference between biomedical problems that are and are not currently susceptible to precise mechanistic analysis.