Persistent systemic inequality leads to negative outcomes for vulnerable groups in many contexts, such as healthcare and employment. Systemic interventions are critical for long-term change, but need complementary approaches that are quick, effective, and accessible to individuals. My research explores how to support marginalized individuals and their allies to take immediate, short-term action to reduce negative outcomes for vulnerable groups.

I approach this problem through the lens of self-efficacy games. Self-efficacy, or the belief that one can accomplish a given task, is a well-known precursor of behavior change; prior work shows that games designed with self-efficacy principles in mind can provide safe environments for players to tackle real-world challenges and increase their self-efficacy with respect to tasks portrayed in the game. However, prior work has also shown that the narrative design of self-efficacy games influences their effectiveness. For example, casting the player as a character who is an expert can negatively affect their risk perception (e.g., in domains such as cybersecurity), one of the four key factors for self-efficacy interventions. My research therefore seeks to better understand how narrative design affects the impact of self-efficacy games.

To accomplish this goal, I have developed self-efficacy games in two domains: health and workplace inclusion. First, I created the visual novel Bloomwood Stories, which is targeted at players who have negative experiences with formal healthcare for reasons such as medical racism or language barriers, and aims to increase their self-efficacy around using informal healthcare resources such as those at their local library. In addition to producing design knowledge, I conducted a quantitative study that found that self-reported player effort predicted self-efficacy gains both immediately after play, and at follow-up one week later. Experience-taking with the non-player characters also predicted self-efficacy gains at post-test, which suggests two different mechanisms for self-efficacy change. To better understand this finding, I conducted a qualitative study to understand how players related to NPCs and the game's health resources, providing insight into the mechanisms that supported their self-efficacy.

In parallel, I collaborated on the development of INTENT, which aims to transform allistic workers into effective advocates for their autistic colleagues. Like Bloomwood Stories, this game is designed with self-efficacy principles in mind, and aims to motivate action by individuals to help mitigate the impact of systemic problems. However, Bloomwood Stories is designed for players who share the challenges of one or more NPCs. In contrast, INTENT is meant for players who do not have lived experience of autism. The NPCs must now serve a double purpose: helping players understand the experiences of their autistic peers, while simultaneously improving player self-efficacy. Given our findings that NPC experience-taking is one mechanism for impact, I conducted a qualitative study with autistic individuals to better understand narrative techniques to responsibly and effectively represent the autistic community.

Taken together, these studies contribute to the process of designing narrative in self-efficacy games that deal with sensitive topics for marginalized communities. By examining the distinct roles that NPCs play across contexts, and identifying mechanisms such as effort and experience-taking, this work advances our understanding of how games can support behavior change through narrative and design.

152 pages

Thesis Committee:
Jessica Hammer (Co-chair)
Geoff Kaufman (Co-chair)
Andrew Begel
Rina Wehbe (Dalhousie University)

Nicolas Christin, Head, Software and Societal Systems Department
Martial Hebert, Dean, School of Computer Science

Return to: SCS Technical Report Collection
School of Computer Science

This page maintained by reports@cs.cmu.edu