Game-Based Learning of Data Structures Based onAnalogies: Learning Gains andIntrinsic Motivation in Higher Education Environments

Abstract

Many researchers have considered video games as effective learning tools. According to them, video games can increase intrinsic motivation and promote active learning. However, video games are a flexible medium, and they allow the integration of many learning theories and pedagogical approaches. This dissertation is concerned with exploring and evaluating the use of video games for learning based on analogies to facilitate learning conceptual knowledge, specifically binary search tree conceptual knowledge. The hypothesis of the thesis is that video games for learning based on analogies are effective tools to teach non-intuitive conceptual knowledge.

To verify our hypothesis, first, a systematic literature review concerning video games for learning that focus on data structures was conducted. Specifically, the review focused on the game elements used to convey learning, the learning aspects of the game, and the evaluation methodologies followed. Second, a review of literature concerning design of learning experiences, entertainment game design, and design for video games for learning was performed. Informed by this review, a game design framework for video games for learning was developed. Third, an action-adventure video game based on analogies that focuses on binary search trees called DS-Hacker was designed and developed using our framework. Fourth, an assessment tool that measures conceptual knowledge about binary search trees was developed and validated. Fifth, two evaluations were performed to verify the effectiveness of our video game as a tool for learning.

Our findings show that video games for learning can teach abstract conceptual knowledge with an efficacy that is comparable to other traditional digital methods, such as video lectures. However, regarding intrinsic motivation, they are more motivating than traditional digital methods. Regarding the pedagogical approach, analogies embedded in the game rules and game mechanics are effective and facilitate the acquisition of new information. Furthermore, the level of realism of video games can slightly affect learning gains. However, this is caused by game elements that increase the external cognitive load.

The contributions of this thesis relate to the knowledge about game-based learning based on analogies that efficiently achieve learning objectives. First, the thesis identifies characteristics, such as strengths and weaknesses, of existing video games for learning that focus on data structures through the systematic literature review. Second, the thesis presents an approach for the design and development of games for learning, which emphasizes the alignment of the context, the learning aspects, and the game design aspects. Third, the thesis presents a validated assessment tool to measure binary search tree conceptual knowledge that generates measures with scale properties that can be used in computer science education research. Finally, an original video game for learning data structures was used to validate the game design framework and to evaluate the efficacy of game-based learning based on analogies.