What is it about?
We produce numbers in all sorts of ways. Some of them are visible, such as a statistical figure (employment rate of a country) or a ranking (which is the N.1 university in the world) of the prediction of a model (we consume 1.7 planets every year). Some are hidden, e.g. the result of an algorithm for conceding credits. Some appear quite remote from the act of quantifying, such as a software for facial recognition. All these instances of quantification are different and yet they share some common pathologies. Which ones? How can these be tackled?
Photo by Josh Riemer on Unsplash
Why is it important?
There is an incredible heterogeneity in the degree of awareness of the problem in the different families of quantifications. While in statistics a 'statistical war' is in full swing, and all problems are brought up in the open, in the field of mathematical modelling everything seems to be going well. While the problems with the use of metrics are debated by experts, society still has a huge appetite for them. Why ethics of algorithms is a present concern, commercial and corporate interests my game the debate in their favor.
Read the Original
This page is a summary of: Ethics of quantification or quantification of ethics?, Futures, December 2019, Elsevier,
You can read the full text:
"Ethics of quantification or quantification of ethics?", Keynote presentation for the Annual symposium of the Centre for the Study of the Sciences and the Humanities (SVT): Ethics of quantification, Bergen, December 6, 2019.
These slides contains the argument and the main sources used for the analysis.
Discussion Paper: Should statistics rescue mathematical modelling?
This is a technical work describing some of the elements of the problem with quantification, specifically the similarities between statistical and mathematical modelling and how these could help to idntify solutions.
A short comment on statistical versus mathematical modelling, Nature Communications, 10, Article number: 3870, https://doi.org/10.1038/s41467-019-11865-8
Short non-technical text on statistical versus mathematical modelling.
The following have contributed to this page