“All Models Are Wrong, Some Are Useful”: George Box
Published version
Peer-reviewed
Repository URI
Repository DOI
Change log
Authors
Routledge, Matthew
Conway Morris, Andrew https://orcid.org/0000-0002-3211-3216
Abstract
In the aftermath of the Covid-19 pandemic, Box’s wisdom regarding models seems as prescient as ever. Much of the more dubious modelling performed during this period has generated substantial criticism.(1) Modelling complex systems involving human beings, incapable of reliably living up to the role of rational actor is particularly fraught with danger.(2) Given the influence of much of this modelling over public health decision-making, it is tempting to add “some are harmful,” to Box’s aphorism. It is therefore, with perhaps some trepidation that we approach the report by Donnelly et al., published in this edition of the Annals of the American Thoracic Society(3). The question follows, is their model useful or potentially harmful?
Description
Keywords
Humans, Sepsis
Journal Title
Annals of the American Thoracic Society
Conference Name
Journal ISSN
1546-3222
2325-6621
2325-6621
Volume Title
Publisher
American Thoracic Society
Publisher DOI
Sponsorship
MRC (MR/V006118/1)
ACM is supported by a Clinician Scientist Fellowship from the Medical Research Council (MR/V006118/1)