Students’ Construction of Science Simulations: “Is That Real Enough?”

• In Event: Examining Computational Thinking in the Field

Abstract

Having students create and consider models of scientific phenomena is an key way to develop their understanding of science. Creating computer models or simulations of science phenomena is a particularly powerful approach (Resnick & Wilinsky, 1999). Not only do students apply their ideas to create the simulation but they need to use computational thinking in the context of thinking about science. Computational thinking can be a powerful tool for thinking about science and many other types of problems (Wing, 2006). Creating computer models is a challenging task for students, but new tools such as NetLogo, STARLogo and Scratch have made this type of activity more accessible to students than ever before (Klopfer & Scheintaub, 2008).

This paper looks at students developing simulations of physics experiments as part of a class on game development for middle school students. In this class, students are developing their computational thinking skills as well as their science knowledge. The use Scratch to create their own versions of popular games (e.g. Angry Birds). The paper presents a video analysis of three groups of students working on a simulation of a ball dropping. Students in the video are tasked with making the fall look realistic. The 90-minute videos were transcribed and coded for discussion of computational thinking and for discussions of science and realism of the resulting motion.

As novice programmers, the students did not have the computational thinking skills to be able to put their ideas of the science directly into the simulation. Instead they made multiple attempts using increasingly more complex algorithms as the demand for realism pushes them beyond their comfort level. Discussion in the groups focused on identifying new approaches for making the ball move in a trial and error approach rather than working from scientific principles. The students made multiple attempts at coding the motion of the ball before deciding that their fall is accurate enough. Social influence plays a significant role as one group strove to be the first in the class to solve the problem, while the others looked for clues in what their peers were doing. Students appeared to gain confidence when they see the results they expect and become eager to use the new tools that they master. Their success in developing their program is aided by key features in the programming interface such as color and shape coding or instructions with different functions.

These cases illustrate the challenges in using scientific modelling and promoting computational thinking. Experience working through problems such as the ball drop suggest that students can make gains in computational thinking in a relatively short time when motivated. Finding motivations that work for all students is a key challenge for promoting computational thinking widely.

Author

• Brian J. Foley, California State University - Northridge