What is it about?
The crisis in statistics — including calls for scientists to abandon statistical significance — should inspire a similar re-think of mathematical modelling,
Featured Image
Why is it important?
Unlike statistics, mathematical modelling is not a discipline. It cannot discuss possible fixes in disciplinary fora under the supervision of recognised leaders. It cannot issue authoritative statements of concern from relevant institutions such as e.g., the American Statistical Association or the columns of Nature. Hence existing problems risk remaining unaddressed. See e.g. some of the problems here: https://growkudos.com/publications/10.1016%25252Fj.envsoft.2019.01.012/reader
Perspectives
Mathematical modelling could benefit from structure and standards based on statistical principles including a systemic appraisal of model uncertainties and parametric sensitivities. Statistics could help by internalising these into its own syllabi and practices. Perhaps what is needed is a new ethics of quantification. An oncoming symposium at the University of Bergen is addressing the theme, https://www.uib.no/en/svt/127044/ethics-quantification.
Professor Andrea Saltelli
University Pompeo Fabra, Barcelona School of Management
Read the Original
This page is a summary of: A short comment on statistical versus mathematical modelling, Nature Communications, August 2019, Springer Science + Business Media,
DOI: 10.1038/s41467-019-11865-8.
You can read the full text:
Resources
Discussion Paper: Should statistics rescue mathematical modelling?
Extended text which elaborates on the short comment providing references, examples, etc.
Ethics of quantification; Annual symposium of the Centre for the Study of the Sciences and the Humanities (SVT)
The intermeshing of algorithms and big data increasingly blurs many existing distinctions between different forms of quantification and is gaining prominence in the nascent field of the sociology of quantification. As Elizabeth Popp Berman and Daniel Hirschman enquire: “What qualities are specific to rankings, or indicators, or models, or algorithms?” In recent times, examples of quantification have come under intense scrutiny with regard to perceived misuse or abuse of existing methodologies. Algorithms pose the risk of non-transparent, oft-proprietary tools being used to decide upon everything from convicting felons to conceding mortgages, and from recruitment to adoption. Quantification as a policy support tool is also found wanting. The convulsions of significance testing in statistics have received wide attention in both media and academia. Mathematical modelling is a field where severe problems are not yet fully charted. Common to many forms of quantification is a singular conundrum: any number which does not represent its context and purpose of production can obfuscate as much as it can illuminate. Is time ripe for an ethics of quantification? The symposium on 6th of December 2019 will feature keynotes by prominent scholars of Science and Technology Studies, the politics of metrics, and the ethics, fungibility and policy relevance of quantification. We are delighted to welcome professor in History of Science, Theodore Porter,UCLA, Professor emerita Catherine Paradeise, Université Paris-Est, Ana Viseu, Associate Professor at Universidade Europeia together with SVTs Andrea Saltelli as our distinguished keynote speakers. The closing round table of the symposium will be introduced and chaired by SVTs Jeroen van der Sluijs. The day prior to the symposium, December 5th, there will be a workshop on the same themes, see https://www.uib.no/en/svt/127044/ethics-quantification.
Ethics of quantification, a lecture given at the University of Bergen, August 2019
A presentation on the topic of ethics of quantification
Why so many published sensitivity analyses are false: a systematic review of sensitivity analysis practices, Environmental Modelling and Software, Volume 114, April 2019, Pages 29-39.
Related article showing how good practices are not always taken in modelling studies.
Contributors
The following have contributed to this page
