In Models We Trust—But First, Validate
By Daniel Solow
Scientists use mathematical descriptions and simulations to make forecasts and provide recommendations to decision makers, but how reliable are these models?
This Article From Issue
January-February 2023
Volume 111, Number 1
Page 42
All models are beautiful—even mathematical ones. They provide insights to scientists and to the general public about a vast array of phenomena in the real world, and we have come to rely on them much more than you may realize.
QUICK TAKE
- Mathematical models inform our understanding of the world. With varying degrees of success, they predict storm tracks, election results, and the motion of falling objects.
- No model is perfect. Researchers who use mathematical models should question how they were validated, and what assumptions lie behind them.
- Common validation methods include commonsense reality checks, comparison to experimental data, updating of parameters, confidence intervals, and sensitivity analysis.
American Scientist Comments and Discussion
To discuss our articles or comment on them, please share them and tag American Scientist on social media platforms. Here are links to our profiles on Twitter, Facebook, and LinkedIn.
If we re-share your post, we will moderate comments/discussion following our comments policy.