Skip Over Navigation Links

Web Exclusives: Diseases

10 Reasons to Model
By Emily Carlson
Posted January 7, 2009

Joshua Epstein talked about these and other reasons for modeling during a recent meeting.
Joshua Epstein talked about these and other reasons for modeling during a recent meeting. Watch his presentation, which begins at 00:12:18.

While you were busy making holiday shopping lists and checking them twice, we at Computing Life were busy compiling another list: reasons to develop computational or mathematical models.

The list, presented in no particular order, comes to you courtesy of computational modeler Joshua Epstein of the Brookings Institution. Epstein pioneered methods to simulate and study how individuals' actions impact an entire community. This list includes just 10 of the reasons Epstein gave to anthropologists, psychologists, neurobiologists, and others who met recently to talk about modeling social behavior.

Read about scientists who model for these very reasons by following the links.

Exclamation iconRead about more reasons to model in Epstein's "Why Model?" External link article.
  1. Explain. You could design the perfect experiment, but the results might leave you baffled. Models can reveal phenomena not easily observed in the lab or the real world.
  2. Predict. To a certain extent, models can predict possible outcomes—global warming, hurricanes, loss of biodiversity, spread of disease.
  3. Discover new questions. A good model can yield unexpected ideas and relationships, leading us to new hypotheses, new experiments, and new findings.
  4. Illuminate abstractions. Real life is so complex and complicated that models, at best, can represent only a simplified and incomplete version of reality. But just like a snapshot, they can reveal part of the bigger picture—and even show what's missing.
  5. Suggest analogies. Could a revolution spread through a community like an epidemic? Are cells organized like the Internet? Modeling can show that two seemingly unrelated events or systems could be driven by the same underlying process. If that process is well studied, researchers can use the information to answer the question at hand.
  6. Demonstrate tradeoffs or suggest efficiencies. Models are like stovetops with different knobs that can adjust the heat, speeding up or slowing down outcomes. By turning the knob a little more or less, researchers can see what variables greatly change the outcome. In a model of drug-resistant bacteria, researchers can "turn the knob" to see how they might slow or even stop the development and spread of such resistance.
  7. Direct policy. A model can serve as a tool for policymakers, who might use modeling outcomes to guide the development of planning strategies for disease spread, natural disasters, and even traffic congestion.
  8. Educate and train. Public health and disaster relief workers can use models in training exercises to prepare for possible real-life events. Models can also show college biology majors the value of mathematics and computational science in studying health and help engineering students envision careers in biomedicine.
  9. Reveal the simple to be complex and vice versa. Modeling can show that a complex question may have a very simple answer. Studies of bird flocking and bee swarming, for example, show that simple rules can direct a seemingly complex behavior. Modeling can also reveal complexity in simple systems, like the structure of a virus.
  10. Promote a scientific habit of mind. Designing a model forces you to think logically and clearly. You have to consider what you know, what you don't know, and what data you have available.

Learn about related research

This page last reviewed on April 22, 2011