Whitkin, Josh (2013). “An Investigation of Learning Game Design: Scoring Activity-Goal Alignment”. Ph.D. dissertation. School of Design, Curtin University, 2013.
This research aims to improve the practice of designing educational video games (“learning games”). This thesis focuses on Shelton’s theory of activity-goal alignment, which highlights the relationship between a player’s activity and the designer’s intended learning goal in any learning game.
The thesis has two sections. First, it extends Sheldon’s theory to be more practical for learning game designers. It develops a tool for scoring alignment, and evaluates it through analytic discussions of designs of several existing learning games. It finds that scoring activity-goal alignment is more useful than Shelton’s narrative-based method, in commercial design studio context.
Secondly, this thesis argues that there is need for improved tools for assessment of learning game designs. It reviews existing assessment tools to develop criteria for any learning game assessment tool. It operationalizes three theories (Higher Order Learning theory, Gee’s principles of Deep Learning, and Shelton’s activity-goal alignment) to develop a the AGA-Based Assessment Tool. It applies the tool in critical discussions of several learning games to reveal important gaps between learning game design practice and theory.
The thesis concludes that scoring activity-goal alignment can be useful to the learning game designer because it makes an important theoretical position from the learning game design literature clear and simpler to apply in practice.
The Essentials of Activity-Goal Alignment Theory
Essentially, activity-goal alignment theory considers the alignment between the player’s activity (e.g. driving, flying, shooting) and the learning goal (e.g. arithmetic, systems thinking, history). Consider its three component words: activity, goal, and alignment.
- Activity refers to video game player’s action (e.g. driving, flying, shooting, and puzzle-solving) within the game. All video games have some kind of activity.
- Goal refers to an educational purpose (e.g. arithmetic, systems thinking, history). All learning games have a learning goal.
- Alignment can be seen in the Venn diagram above. The degree of overlap of the two circles illustrates how related the two entities are.
For example, imagine an instructional driving game, where the player learns to drive a car by playing a 3D simulation of normal driving on public roads. Such a game has high activity-goal alignment because the activity of driving the car in the game resembles a real-world driving experience. By contrast, a road safety quiz has low activity-goal alignment, because the activity of taking a test and recalling facts is not very similar to driving a car.
- An Introduction for Commercial Designers
- an interactive tool for thinking visually about activity-goal alignment in your project
Dissertation Chapter Summary
The dissertation has eight chapters:
- Literature Review
- Scoring Activity-Goal Alignment
- Using Activity-Goal Alignment Scores in Analysis of Games
- Review of Learning Game Design Assessment
- Activity-Goal Alignment-Based Assessment of Learning Games
In the introductory chapter I set out the basic ideas, establish the need and significance of this research, define terms and outline the limitations, methods, and structure of the dissertation.
The second chapter is the literature review. I present Shelton’s original conception of activity-goal alignment theory, and relate it to established learning game design theory (e.g. fidelity, mechanics, alignment) as well as important ideas from the fields of education, training simulation, and commercial video game design.
In the third chapter I present the method for this thesis. I begin by describing my practical activities within this research, and then present my rationale. I fit this research in the ontology of research methodology, and argue for a design-based approach. I argue for ethnography, and specifically analytic autoethnography, as the best fit for the constraints and goals of this research. I then discuss known weaknesses of the method, and address ethical considerations.
In chapter 4 I establish the need for scoring activity-goal alignment, and describe my development of the AGA Scoring Tool to score activity-goal alignment. I start with an autoethnographic narrative describing my attempt to achieve extremely high activity-goal alignment in the development of Battlefood, a learning game. I discuss examples where activity-goal alignment would have been useful, and note that Shelton’s application of activity-goal alignment theory could be amended to better suit practicing designers’ needs. In accordance with my method, I expose my process and explain my design reasoning in my creation of the AGA Scoring Tool.
In chapter 5, I exercise the AGA Scoring Tool by using it to analyze five different learning games. I draw from a variety of data sources including prior analyses, my observations of others’ play experiences, and my own experience. I discuss a broad range of games, including the well-regarded Re-Mission and GameStar Mechanic as well as more obscure and unsuccessful learning games. In the first section I discuss several games’ activity-goal alignment using a discussion based on Shelton’s approach, and then score their activity-goal alignment using the tool built in chapter 4. I find that scoring activity-goal alignment is useful. In the second section, I probe for weakness in both activity-goal alignment theory and the scoring tool. In the end, I warn designers that scoring activity-goal alignment does not shortcut the reflection-in-action of good design practice. I find that multiple, diverse activities are more difficult to represent clearly in the activity-goal alignment scoring diagram, and that scoring activity-goal alignment illuminates differences between two games about a single learning topic. Finally, I discuss practical issues I encountered scoring activity-goal alignment.
In chapter 6, I explore how activity-goal alignment might be used for assessment of quality, not just design investigation, in learning games. I discuss existing methods of assessment of learning games, from both commercial and academic sources. I then create a second new tool, the AGA-Based Assessment Tool. This tool combines activity-goal alignment and Higher Order Learning theory to create four categories of learning games. I discuss how this tool’s unique approach could benefit learning game designer/researchers.
In chapter 7, I exercise the AGA-Based Assessment Tool developed in chapter 6 to categorize several learning games. I discuss examples that fail to respond to key advice from veteran learning game designers, as well as exemplary games. I relate the tool’s output to important design advice in the literature, and find that the tool is useful within limitations. I argue that activity-goal alignment-based categorization guides designers towards better learning game design because the results are difficult to misinterpret, advising designers to avoid one type of game design, and target another.
In chapter 8, the conclusion, I discuss the implications of this research. I discuss how findings from chapters 4 and 5 suggest that scoring activity-goal alignment is a useful “rule of thumb” that is quicker and more accurate than textual discussion of activity-goal alignment, although it does have drawbacks. In discussing chapters 6 and 7, I first note that the literature effectively advises designers to aim for high activity-goal alignment, but low-activity-goal alignment games are still being built despite this advice. I argue against using low-activity-goal alignment game designs in research, and suggest that scoring activity-goal alignment might help designers stay focused on the important opportunities in learning game innovation. I discuss how this work contributes to the literature, focusing on Shelton’s original conception of activity-goal alignment. Finally, I consider directions for further research.
Full text of dissertation as PDF available upon request. email me here.