1 September 2010
Before starting grad school, I worked for Peace Corps Mongolia as a community development specialist on projects such as starting a small dairy business with a local school and starting a nation-wide network of after school clubs to teach life skills (communication, decision-making, and critical thinking). After returning to the U.S., I got involved with teaching a class at Carnegie Mellon University called Technology-Consulting in the Community a kind of "Peace Corps for Geeks" that trains computer science students how to work with local and international non-profits to develop technology projects. These experiences got me interested in policy problems: how do we prevent nuclear proliferation, global warming or poverty?, or at the local level: how does our village decrease the rate of sexually-transmitted infections, or increase economic activity?
Unfortunately, students face a number of challenges when learning how to solve policy problems, for example, Technology Consulting students have difficulty choosing projects that directly impact the mission of the non-profit partners. Furthermore, whether it's Peace Corps volunteers, technology students, public policy professionals, or simply citizens trying to vote–people have difficulty making policy decisions on the basis of evidence. Much of my graduate work consists of empirical studies showing that when given evidence that contradicts one's policy beliefs, people tend to ignore the evidence rather than change their mind.
At Carnegie Mellon University's Human-Computer Interaction Institute, we design and study systems that bring humans and technology together to achieve a purpose. In my case, that means using cognitive tutors (a kind of artificial intelligence applied to education) to help citizens and students learn how to solve policy problems. For my dissertation I designed a cognitive game called Policy World that combines an educational game and cognitive tutor to teach students how to use diagrams to solve policy problems.
Policy World is a Phoenix Wright-inspired adventure game. Students play a policy analyst working for a think tank that must make recommendations to a senator about policy issues like lowering the legal 21 drinking age, health care, video game violence, junk food advertising, cap and trade, methamphetamine epidemic. In the first part of each level, the student uses a (fake) Google interface to search for newspaper articles, scientific reports and editorials about a policy topic. Along the way the student uses an interactive diagramming tool to represent the evidence (Figure 1) that has been shown to improve students' ability to make evidence-based recommendations. At the end of each level, the student has to debate an opponent by using evidence to convince the senator about which policy to implement (Figure 2).
Policy World was developed with Adobe's Flash Builder, Creative Suite, the Mate Flex Framework, and the Degrafa graphics framework. Policy World can be deployed as an AIR application for instructors interested in a stand-alone version or run as a web-based application deployed on the Open Learning Initiative on-line course delivery platform.
This allows Policy World to collect extensive information about students' problem-solving processes. This information provides feedback to instructors and designers that is used to improve the effectiveness of the game. In an experiment testing different versions of the Policy World, one with only "traditional" game feedback and another version with a cognitive tutor, we found that adding a cognitive tutor to the game increased both learning and motivation. So embedding cognitive tutors in games proves to be an effective way to use educational technology to create engaging and effective learning experiences.
Now that Policy World has progressed past the proof of concept phase, we are adapting the game for use in public policy classes. We see this as a first step toward using educational technology to support policy argument education across a wide range of curriculum that help students, citizens, and organizers to make evidence-based policy decisions.
Instructions for logging onto the Policy World Demo: www.matteasterday.com/policyworld.html
Technology Consulting in the Community: www.cs.cmu.edu/tcingc/
The Open Learning Initiative: oli.web.cmu.edu/openlearning/
Human-Computer Interaction Institute: www.hcii.cmu.edu
Tutorials & Samples