Duncan Williams
· ProfessorVerifiedUniversity of Southern California · Religion
Active 1986–2026
Research topics
- Computer Science
- Psychology
- Social psychology
- Computer Security
- Political Science
- Artificial Intelligence
- Business
- Cognitive psychology
- Sociology
- Engineering
- Internet privacy
- Law
- Art
- Finance
- Literature
- Visual arts
- Human–computer interaction
- Commerce
Selected publications
Social Science Computer Review · 2026-01-19
articleSenior authorThis study investigates networked social influence in Sky: Children of the Light , a social multiplayer online game. Drawing on survey responses ( n = 9,254) and in-game data from over 660,000 players, we use an innovative graph-based machine learning approach to quantify how individuals influence others’ playtime, and regression analyses to test predictors from the COM-B model. Results show that Capability enhances influence, although excessive task focus correlates negatively with social impact; Opportunity emerges as the strongest predictor, with active social interactions significantly boosting influence; and Motivation varies by playstyle, with socializers and competitors demonstrating greater influence than narrative-focused players. By applying the COM-B model in a digital gaming context, this research highlights behavioral dimensions of player influence and employs a novel metric for quantifying interpersonal influence. These findings suggest practical implications for game design, particularly by highlighting how social interaction opportunities and different player motivations shape influence within communities.
Nonprofit Management and Leadership · 2026-04-11
articleSenior authorABSTRACT Although previous studies tend to emphasize the role of influencers in donation campaigns, this study shifts attention to focus on influencees, namely donors who are highly receptive to the influence of others within local social networks. We integrated insights from social influence research with that of the COM‐B model and examined the distribution of influencees in eight donor populations. In addition, we examined how donors' capabilities, opportunities, and motivations influence their propensity to be influencees. We found that across eight mutually exclusive donor networks, about 10% of donors can be classified as influencees and they, on average, donate two to 10 times more than other donors. Our analysis consistently demonstrated that donors with extensive connections and those associated with tightly interconnected clusters were more inclined to be influencees. Social influence was clearly present and played a large role for these influencees. Additionally, both the capability and motivation of donors played crucial roles in accurately pinpointing the most receptive individuals to social influence. Notably, disparities were observed among different racial groups. Overall, models that integrated all three sets of COM‐B factors exhibited superior predictive performance.
Change is Hard: Consistent Player Behavior Across Games with Conflicting Incentives
ArXiv.org · 2026-03-17
articleOpen accessThis paper examines how player flexibility -- a player's willingness to engage in a breadth of options or specialize -- manifests across two gaming environments: League of Legends (League) and Teamfight Tactics (TFT). We analyze the gameplay decisions of 4,830 players who have played at least 50 competitive games in both titles and explore cross-game dynamics of behavior retention and consistency. Our work introduces a novel cross-game analysis that tracks the same players' behavior across two different environments, reducing self-selection bias. Our findings reveal that while games incentivize different behaviors (specialization in League versus flexibility in TFT) for performance-based success, players exhibit consistent behavior across platforms. This study contributes to long-standing debate about agency versus structure, showing individual agency may be more predictive of cross-platform behavior than game-imposed structure in competitive settings. These insights offer implications for game developers, designers and researchers interested in building systems to promote behavior change.
Understanding human behaviour for pandemic preparedness with epigames
Nature Health · 2026-02-24
articleChange is Hard: Consistent Player Behavior Across Games with Conflicting Incentives
arXiv (Cornell University) · 2026-03-17
preprintOpen accessThis paper examines how player flexibility -- a player's willingness to engage in a breadth of options or specialize -- manifests across two gaming environments: League of Legends (League) and Teamfight Tactics (TFT). We analyze the gameplay decisions of 4,830 players who have played at least 50 competitive games in both titles and explore cross-game dynamics of behavior retention and consistency. Our work introduces a novel cross-game analysis that tracks the same players' behavior across two different environments, reducing self-selection bias. Our findings reveal that while games incentivize different behaviors (specialization in League versus flexibility in TFT) for performance-based success, players exhibit consistent behavior across platforms. This study contributes to long-standing debate about agency versus structure, showing individual agency may be more predictive of cross-platform behavior than game-imposed structure in competitive settings. These insights offer implications for game developers, designers and researchers interested in building systems to promote behavior change.
medRxiv · 2026-01-16
articleOpen accessAbstract Background Understanding the drivers of protective behavior during infectious disease outbreaks is critical for public health policy. App-based experimental epidemic games (epigames) offer a novel method to study these behaviors empirically, but their external validity—how well in-game choices reflect real-life beliefs—still needs to be rigorously tested. Methods We conducted a two-week randomized controlled trial (N=567) using the Epigames smartphone app at the American University of Iraq – Baghdad (AUIB) campus in the Middle East. This app used Bluetooth communication to sample the contact network between participants in sub-minute resolution and simulated the spread of a hypothetical respiratory virus through this network. Participants were randomized into two groups with differing opportunity costs (in-game points) for adopting voluntary quarantine within the game that would protect them from the simulated infection: Group 1 (Low Barrier) faced a small point difference between quarantine and non-quarantine choices, while Group 2 (High Barrier) faced a much larger difference. The optimal point difference between groups was determined by a Willing to Accept (WTA) pilot survey prior to the epigame. We measured real-life health beliefs and in-game beliefs through surveys at the beginning of the epigame, including questions on susceptibility, severity, self-efficacy, and benefits, to calculate the correlation between the two and to construct a Health Belief Model (HBM) parameterized by game data. We also measured self-assessment of game realism and behavior via an exit survey. Results Real-life and in-game beliefs showed moderate positive (Spearman’s coefficient ρ from 0.13 to 0.36) and statistically significant (p-value < 0.05) correlation across all survey measures, suggesting indicator (behavior) parallelism between real-life and the epigame. The exit survey also yielded positive self-assessment of context and behavior realism during the epigame. While simple aggregate analysis showed no significant difference in quarantine rates between Groups 1 and 2, a Poisson regression model revealed a significant crossover interaction. High economic barriers significantly reduced quarantine adoption among participants with low in-game motivation (interaction coef. = -2.22, p < 0.01). Perceived benefits appeared to moderate this effect at a near significance level (coef. = +0.33, p < 0.08). Demographic factors such as gender appeared to be significantly correlated with quarantine choice (p < 0.01). Analysis of the contact network measured with the app through Bluetooth showed assortative properties of several belief variables. Conclusion This is the first systematic pre-registered study on the external validity of app-based epigames and the impact of economic cost and individual beliefs on protective behaviors. We found that economic barriers act as a “gatekeeper” for quarantine during the game, suppressing action among the skeptical while allowing highly motivated individuals to act. The significant correlation between real-life, in-game beliefs, and network structure suggests that epigames are a valid experimental tool for network-aware behavioral epidemiology. This warrants further studies for replication, examining variance across settings, and addressing limitations (e.g., crosstalk between groups) and technical issues (e.g., Bluetooth connectivity) in the study design. Study pre-registration: https://osf.io/qev6n/
Journal of Media Psychology Theories Methods and Applications · 2026-04-02
articleSenior authorAbstract: While gender and race are known to influence self-presentation in digital environments, how critical socio-political identities such as political ideology, military experience, and national identity shape in-game behaviors remains underexplored. Building upon Social Identity Theory (SIT), this study examines how these identities predict national identification through the selection of country-based vehicle. By integrating unobtrusive behavioral log data from World of Tanks, self-reported survey data, and global in-game statistics, we found that North American and Russian players disproportionately select US and Soviet tanks, respectively. At the player level, American respondents ( N = 2,113) with military experience or conservative ideology show significantly higher selection rates for US vehicles and lower usage of Chinese tanks. Crucially, this study quantifies the psychological friction between expressive identity and instrumental competition motivation. We found that players align with their national identity even when such choices are detrimental to winning a game. Our study identifies a psychological subsidy effect, where identity-congruent selection mitigates the importance of performance, revealing a clear hierarchy of motivations in immersive environments. By demonstrating how offline ideological schemas dictate digital decision-making through nonhumanoid avatars, this research expands the conceptual boundaries of SIT. Implications for game companies are discussed.
Change is Hard: Consistent Player Behavior Across Games with Conflicting Incentives
2026-04-13 · 1 citations
articleOpen accessThis paper examines how player flexibility – a player’s willingness to engage in a breadth of options or specialize – manifests across two gaming environments: League of Legends (League) and Teamfight Tactics (TFT). We analyze the gameplay decisions of 4,830 players who have played at least 50 competitive games in both titles and explore cross-game dynamics of behavior retention and consistency. Our work introduces a novel cross-game analysis that tracks the same players’ behavior across two different environments, reducing self-selection bias. Our findings reveal that while games incentivize different behaviors (specialization in League versus flexibility in TFT) for performance-based success, players exhibit consistent behavior across platforms. This study contributes to long-standing debate about agency versus structure, showing individual agency may be more predictive of cross-platform behavior than game-imposed structure in competitive settings. These insights offer implications for game developers, designers and researchers interested in building systems to promote behavior change.
2025-04-23 · 3 citations
articleUnderstanding Social Interaction and Relationships in and through Online Video Games
Oxford Research Encyclopedia of Communication · 2025-09-16 · 1 citations
reference-entrySenior authorVideo games, which have become the largest medium on the planet, serve as virtual third spaces where interpersonal relationships are extended, modified, or even formed and dissolved. Extant research has identified various affordances of massively multiplayer online games (MMOs, or MMOGs) that facilitate or hinder the development of interpersonal relationships. In-game elements, such as in-game groups (e.g., guilds, clans), communication features, and co-play, allow players to interact with others and may encourage self-disclosure and reciprocity, which are key to relationship building. This leads some players to develop and maintain long-term, deeper connections (i.e., strong ties) that would otherwise remain as loose connections (i.e., weak ties). In addition to game affordances and mechanics, scholars have been interested in how individual-level factors interact with in-game features in shaping gaming-related friendships and their effects. It has been shown, for example, that individual differences—such as demographics and motivations—as well as dyadic-level factors, such as the types of co-players, can produce meaningful differences. In-game relationships can be both positive and negative for players and their surrounding networks. MMO players can acquire both bonding and bridging social capital, as well as enhance their well-being by satisfying relatedness needs. Emerging gaming technologies, such as augmented/virtual reality and game streaming, which enable novel forms of social interactions, confer similar benefits. However, concerns have been raised regarding the potential of online games displacing existing offline relationships and provoking identity-based conflicts based on out-group antagonism. To steer the gaming environment into a positive direction, gaming developers need to have a good set of values that help them build social architectures that foster connections and inclusive gaming environments.
Recent grants
Frequent coauthors
- 27 shared
Rabindra Ratan
Michigan State University
- 19 shared
Cuihua Shen
University of California, Davis
- 17 shared
Noshir Contractor
- 15 shared
Jaideep Srivastava
- 12 shared
Brian Keegan
University of Colorado Boulder
- 9 shared
Muhammad Aurangzeb Ahmad
University of Washington Bothell
- 8 shared
Nicole Martins
Indiana University Bloomington
- 7 shared
Marshall Scott Poole
Labs
DornsifePI
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Duncan Williams
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
- Free to start
- No credit card
- 30-second signup