A data-driven, player-centric approach to designing spatial skill training video games

"Spatial skills are a subset of cognitive skills essential to success in many different STEM fields, including engineering, chemistry, geology, and computer science [1, 2]. Those with low spatial skill often struggle in introductory college-level STEM coursework and drop out of STEM majors. Fortunately, spatial skills are quite malleable. Video games present a particularly promising way of training spatial skills; certain commercial video games, such as Medal of Honor, Portal 2, and Tetris, have been empirically shown to improve players' spatial skills after just a few hours of training [3, 4, 5], and video games provide a motivational advantage over other forms of spatial skill training interventions since they are designed to be fun. However, other commercial games, such as the ""brain-training"" game Lumosity, seem to have no effect on players' spatial skills [5]. It is not clear what makes some games effective and others ineffective at spatial skill training, which makes it difficult to design game-based spatial skill training interventions to improve students' STEM proficiency and retention. Prior work studying the effectiveness of game-based spatial skill training interventions is also limited by the fact that it does not take the their motivational appeal to their target audience into account - the main advantage games have over other kinds of training interventions. Low spatial skill students, who are disproportionately female, stand the most to gain from spatial skill training interventions through improved proficiency in STEM coursework, but are not targeted in the design or evaluation of spatial skill training games. In this dissertation, I present a data-driven, player-centric approach to designing spatial skill training video games that contributes to our understanding of what game features may contribute to a game's effectiveness at training spatial skills and its motivational appeal to its critical target audience of low spatial skill students. First, I explain the design of Homeworld Bound, a game I designed as a testbed for evaluating the ability of different spatial game features to tap into players' spatial skills and demonstrate its effectiveness as a training intervention for children. In my first study, I demonstrate my data-driven approach to evaluating the effectiveness of specific game features at tapping into players' spatial skills by analyzing the relationship between player performance and spatial skill in Homeworld Bound. My results reveal that most of the levels in Homeworld Bound successfully tap into players' spatial skills and provide insights about how to fix the levels that do not. In my second study, I take a player-centric approach to designing the player experience of spatial skill training games, investigating how demographic factors and gaming habits, preferences, and motivations predict spatial skill in young adults. I use my findings to develop a set of recommendations for designing spatial skill training games that appeal to low spatial skill young adults specifically. I then present a revised and improved version of Homeworld Bound, Homeworld Bound: Redux, and demonstrate how I incorporated my findings from my data-driven and player-centered research studies to improve the game's ability to tap into players' spatial skills and its motivational appeal for low spatial skill young adults. Finally, I present the results of a controlled training study I conducted in a large introductory STEM course for non-majors to evaluate the training effectiveness and motivational appeal of Homeworld Bound: Redux for low spatial skill college students. While I found no training effects of Homeworld Bound: Redux compared to alternative training programs after 70 minutes of training, I found that performance on certain levels of the game was correlated with spatial skill and that low spatial skill students were more intrinsically motivated to play the game than to complete a non-game spatial skill training intervention. My findings in this dissertation contribute a deeper understanding of what features influence a game's effectiveness at training spatial skills and motivating low spatial skill students to play it. These findings can be used to inform the design of more effective and motivating spatial skill training interventions in the future to promote the development of skills critical for pursuing STEM majors and careers."

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Computer Science

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2020

Author dc:creator

• Wauck, Helen Catherine

Contributors dc:contributor

• Bailey, Brian
• Zilles, Craig
• Herman, Geoffrey
• Lane, Chad
• Lucas, Gale

• spatial skills
• spatial reasoning
• STEM education
• cognitive training
• video games
• player experience
• educational games

Statement dc:rights

• Copyright 2020 Helen Wauck

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/107962

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/107962