Neo-Piagetian Classification of Reasoning Ability and Mental Simulation in Microsoft's Kodu Game Lab

Over the past five years, there has been a major push to develop the computational thinking skills of K-12 students. Tools such as Scratch, Alice, and Kodu have been developed to engage students in learning to program through the creation of computational artifacts (e.g., games, animations, and stories). However, less is known about how elementary and middle school children reason about program behavior. Such skills are useful for reading and adapting others programs, locating possible sources of bugs, and predicting program behavior given code snippets (i.e., mental simulation). The goal of this poster is to measure and track the development of students' ability to reason about programs using Teague & Lister's Neo-Piagetian classification of novice programmers: Sensorimotor, Preoperational Thinkers, and Concrete Operational Thinkers. We operationalize Teague and Lister's category descriptions by creating a criterion for each category. This classification has helped us characterize students' mastery of strategies for reasoning about the lawful behavior of programs using a Kodu curriculum. In particular, this categorization was used to differentiate students' reasoning styles using data from two studies having 20 and 19 students each. We found strong consistency in the results across both studies. Through analysis and categorization of student responses, most students fall into the preoperational thinker category. Within this category, we found a diversity of mastery patterns that help us understand where students face challenges in reasoning about programs.