2016 ASEE Annual Conference & Exposition

We believe that educating problem solvers is one of the most critical functions of higher education. In the problem-solving literature, problems are understood in part by the degree to which they are structured. Well-structured problems have clear goals which can be obtained through repeatable algorithms to get absolute correct solutions. In contrast, ill-structured problems are inherently vague, can have several viable solution paths, and solving the problem largely requires subjective arguments that frame and guide the attempted solutions.

The central aim of this paper is to explore the feasibility of using an educational computer game as a novel means of assessing problem-solving competency. In this case, the novelty of the approach compared to traditional methods is twofold: (1) the use of a computer game enables us to directly observe problem-solving process through action in the game and (2) because of the naturally immersive game environment, we hope to see motivation and persistence in the face of complex problems that might otherwise be difficult to achieve in a laboratory setting. As a feasibility study, we present methods and data that we suggest should guide further research but should not be used to make specific claims.

In particular, this project uses the educational game Contraption Maker as a problem-solving research tool. Contraption Maker is a commercial (free for educational use) game involving Rube Goldberg inspired puzzles using simple physics-based interactions among objects. The simple premise of the game is seen as an advantage because it minimizes the importance of domain specific knowledge. This project poses a series of increasingly difficult puzzles to participants. Video recordings of the computer screen allow study of the diverse ways in which participants solve puzzles and retrospective interviews of participants allow us to understand the problem solving strategies used by individuals while attempting solutions.

The paper explains the schemes used to analyze the data along with initial findings from the pilot study with university students. We hope this work will generate conversation about novel ways in which educational computer games can be used as informative research and assessment tools. In particular, because Contraption Maker is being used in some K-12 classrooms as a learning tool, this research could contribute to more robust ways of studying game-based learning. Finally, this work also serves as an important first step towards further utilizing computers in problem-solving research. As we better understand how to port problem-solving tasks in a digital environment like this, we can more fully utilize the capabilities of the technology to automate some data processing. One such example could involve using automatically collected mouse-over data in order to identify how the problem-solving process proceeded.