Inquiry Magazine Northwestern School of Education and Social Policy

FALL 2013

Models using his NetLogo software teach chemistry, biology, climate science, mathematics and more.

Professor Uri Wilensky extends worldwide the use of computer modeling to deepen learning. Models using his NetLogo software teach chemistry, biology, climate science, mathematics and more.

Improving Education Worldwide with Computer Modeling

By Rebekah Snyder
computer modeling

Rapid technological advances are revolutionizing how science, mathematics and complex theories are taught around the world, according to Uri Wilensky, professor of learning sciences and computer science in the School of Education and Social Policy, McCormick School of Engineering and Northwestern’s Institute on Complex Systems. The software he has developed over the past 25 years is now used on every continent in education and research and is on a trajectory to be a primary teaching tool for elementary through university classrooms around the globe.

“When we had the Roman numeral system, it was very difficult to multiply and divide numbers,” says Wilensky. “With the transition to the Hindu-Arabic numeral system, we had an appropriate representation to understand and express concepts such as multiplication and division. It was a revolution in how we worked in mathematics, and more importantly it made areas of mathematics, science and commerce accessible to the public.” NetLogo, the open-source agent-based modeling software that Wilensky first released in 1999, is having the same revolutionary effect today on our ability to teach and understand complex systems. At the same time, increased computing power, processing speed and network technology make it possible to share this powerful tool globally.

Wilensky is founder and director of the Center for Connected Learning and Computer-Based Modeling (CCL), a research group dedicated to “the creative use of technology to deepen learning” in both formal and informal settings, for learners of all ages. At the heart of the CCL’s work, NetLogo has more than 100,000 downloads each year. The software allows users to tap into an extensive library of models to understand complex natural and social phenomena.

Users can also create their own models to control how each “agent” in a complex system behaves individually and how the sum of each agent’s activities impacts the system. Biology, chemistry, physics, psychology, economics, sociology, political science, chemical and industrial engineering: agent-based modeling can improve understanding across all disciplines, says Wilensky.

A Global Education Tool

Since NetLogo was introduced in 1999, Wilensky has expanded its scope and utility through international collaborations. In partnership with the CCL, Oxford University has developed software for embedding NetLogo into its online system, and educators have developed libraries of agent behaviors, which are used in teaching biology, chemistry and social sciences. Similarly, educators at the University of London and the University of California, Berkeley are using the NetLogo environment to support mathematics instruction at the middle and high school levels, developing new libraries of models related to probability.

Agent-based modeling is an ideal teaching tool for climate change science, a growing academic discipline that employs complex systems theory. With a high percentage of the population along the coasts, Australians are particularly threatened by the impacts of global warming. CCL collaborators at the University of Sydney have developed a national global warming curriculum, with NetLogo at its core. “Middle and high school students can model changes in the carbon dioxide cycle and resulting impacts on plant growth, animal populations, ocean levels,” explains Wilensky. Researchers, in turn, can understand how agent-based models can help youth better understand one of the most complex scientific issues of our time.

With researchers at University of Haifa in Israel, the CCL is understanding how to teach chemistry better. The result is a “Connected Chemistry” curriculum: an array of NetLogo models and other learning tools that allow students to control their own experiments. Students “play” in an open simulation, observing how particles respond as they reduce the volume of a container of gas or manipulate temperature in a chemical reaction. They can also connect real-world objects, such as a syringe, to the simulation and use “bifocal” modeling to compare real-world and simulated experiments. “The NetLogo environment allows students to connect their real-world action — such as reducing the volume of a container of gas — to the resulting chemistry phenomenon,” says Wilensky. In this way the curriculum effectively teaches both concepts in chemistry and the fundamentals of the scientific process.

Advancing Research for Improved Policy

Wilensky sees the social sciences as a key area of growth for NetLogo. He believes the disciplines of political science, history, economics, sociology and linguistics can benefit from the analysis of alternative scenarios that is possible through modeling. “When you can model a phenomenon visually, complex systems become easier to understand,” he explains. “Using agent-based modeling to illustrate the dynamics of systems such as evolution, voting behavior or segregation — there isn’t much that can’t be understood in this way.”

Through numerous collaborations, NetLogo is being used to understand complex social issues and inform policy around the world. CCL and the Johns Hopkins University are working with Chinese universities to use NetLogo to explore impacts of Chinese emigration. As the world’s most populous country experiences a dramatic population shift from rural to urban areas, NetLogo is using Chinese data sets to model shortand long-term impacts on congestion, pollution and the economy.

Similarly, a CCL graduate student is using GIS data from the Chicago region to model the impacts of road pricing on congestion. “Some research suggests that a driving tax will reduce congestion, as people choose alternatives to paying,” Wilensky explains. “However, the models show a more complex behavior, with feedback cycles in which congestion first decreases, only to increase again as drivers become willing to pay to drive the less congested road.” Because the models are implemented with the easy-to-read NetLogo modeling language, it is possible to explore the assumptions in this research even for people without modeling experience or high-level math skills. This research is contributing to policy discussions in the Chicago region and has implications for cities around the world.

Bringing Innovation Online

NetLogo is a free open-source application, which is downloaded to the user’s local computer or network. “The future of NetLogo is web-based,” suggests Wilensky. Hosting the program remotely will allow for significant scaling up of usage, both in the United States and globally. Wilensky has received funding from the National Science Foundation to port NetLogo to a browser-based platform, which will make it more accessible on devices such as phones and tablet computers.

The CCL research group is also making agent-based modeling more effective with a networked learning tool, HubNet, built into the NetLogo system. HubNet enables educators to teach about complex systems through “participatory simulations” in which each student is connected through the network to act as an “agent” in a system, interacting with other agents and collaborating on experiments. The result is deep, experiential learning in such areas as evolutionary biology, chemistry, electricity and population dynamics. Together, these innovations create the opportunity for users from all over the world to participate simultaneously in a simulation through their handheld devices or computers. That is science on a truly global scale.

What is the potential for NetLogo to transform education around the world? Wilensky believes agent-based modeling simulation is such a powerful educational tool that, with technological advances and through international collaborations, it will one day be a core tool in every curriculum from kindergarten through university, worldwide. “We’ll get there,” he says. CHRISTINE



Midway through a 10-month Fulbright fellowship in South Korea, Christine Yang is probing different angles on how high school students learn mathematics. The doctoral candidate is hoping to better understand mathematical thinking across cultures, especially how a number of mathematical representations can facilitate deeper or richer thinking.

She is examining students’ learning with both “top-down equation models and bottom-up, agentbased models,” the latter using NetLogo. Through one-on-one interviews, Yang discovers how use of the software advances their understanding. “By engaging in a cross-cultural study, I believe we will shed light onto aspects of student thinking and learning that would not otherwise be visible in a localized study context,” she notes.

In the remaining months, Yang will conduct follow-up interviews, visit classrooms and talk to educators to grasp the broader context of mathematics education in Korea. While many international studies focus on achievement data, Yang’s study zeroes in on processes and “how multiple representations help students in different cultures engage in mathematical thinking.” More broadly, she says, “I hope that this study will further intercultural awareness and understanding.”