Computer simulations have become essential to scientific investigation and engineering design, thanks to advances in mathematical modeling, game theory, and computing technology. Simulations now provide an indispensable tool for investigating the properties of natural and built systems in science, engineering, economics, and social science.

Scientists use simulations to make the invisible visible, answer questions, test ideas, and make predictions. Engineers use simulations to test the safety and performance of engineering solutions. Simulations help us understand dynamic, complex systems like weather and climate, rush hour traffic, ecosystem dynamics, group behavior, and much more.

It is easy to understand the appeal of using simulations in our science classes. Simulations are safe and often free; they require limited equipment and can be completed in a short amount of time. Students can easily design their own investigations, decide which variables to manipulate, and run multiple trials quickly. In addition, simulations often include particle-level views and graphical representations of the phenomena being studied. They can also involve game-like environments that are interesting and engaging for students.

A potential downside of computer simulations is that they can give students a misleading view of the nature of scientific inquiry. Simulations generally produce idealized data and “perfect” results. But science itself is a messy process, filled with ambiguous outcomes and inconclusive data that require problem-solving and critical thinking. The best investigations are often the messy ones, designed by students themselves, which must be analyzed and redesigned because of equivocal results. This messiness is at the heart of science.

Simulations are therefore most valuable when teamed with actual physical investigations and their untidy data. For example, students in my chemistry classes use physical equipment to begin a unit on the properties of gases. After investigating the relationship between gas pressure and volume and graphing their laboratory data, students next explore an online simulation that produces idealized data, a molecular level model of the phenomenon, and a graphical analysis that students compare to their own results. Finally, students research the work of Robert Boyle and discover his elegant 1662 experiment, with its meticulous data and graph of the pressure–volume relationship we now identify as Boyle’s Law. In these lessons students develop understanding of the value of simulations as well as an appreciation for the history and nature of science.