Research

Fairness and Equity in Machine Learning Systems for Gambling Harm Prevention

Machine learning models can be used to predict people’s risk for experiencing gambling-related harms based on patterns in their online gambling behaviour. However, the ability of these models to make predictions about people from different sociodemographic backgrounds can vary if there are systematic differences in a group’s gambling behaviour or risk for experiencing harm. The most serious implication of this fact is the possibility that members of certain groups may be less likely to receive harm prevention resources or treatment referrals if the system is allowed to decide who is (and who is not) at-risk.

In this project, I aim to understand sociodemographic biases in existing machine learning systems used to detect at-risk online gamblers. I am trying to learn how these biases relate to the incidence of gambling-related harm in different groups. In doing so, I am working to develop strategies for making these systems fairer without decreasing their sensitivity to detect harm, and investigating methods for calibrating their decision-making processes based on different groups’ risk for experiencing gambling-related harms.

Additionally, I am working to understand the perspectives of end users of the system, as well as experts in gambling regulation, artificial intelligence, ethics, sociology, and the social services sector. Their input will be used to guide the final design of our harm detection models, ideally resulting in a fairer and more-equitable system ready for deployment in Canada.

Funding Sources:

Canadian Institutes of Health Research, Social Sciences and Humanities Research Council of Canada

Related Publications

Murch, W.S., French, M., & Kairouz, S. (submitted). Online Gamblers’ Preferences for Performance and Fairness in Artificial Intelligence Systems for Gambling Harm Detection.

Murch, W.S., French, M., & Kairouz, S. (2024). Comparing ‘Fair’ Machine Learning Models for Detecting At-Risk Online Gamblers. International Gambling Studies. doi: 10.1080/14459795.2024.2412051. [link] [preregistration] [open materials]

Development of a Detection System for Problematic Online Gambling

Spurred by recent advances in internet connectivity and mobile computing, and compounded by the COVID-19 pandemic, an increasing number of people are participating in online gambling. This presents new opportunities for the field of Responsible Gambling (RG). The requirement of online gamblers to maintain unique user accounts enables gambling operators to deliver targeted interventions that may improve the efficacy of specific RG initiatives. However, the tenability of targeted interventions requires accurate identification of users experiencing gambling-related harms.

In this project, I am working with a talented team of sociologists, statisticians, and econometricians to develop machine learning algorithms that use online gambling behaviour to identify people who are experiencing moderate-to-severe gambling problems. These algorithms may be used to modify the normal functioning of online gambling platforms, curtailing the number of promotions and advertisements delivered to individuals experiencing gambling-related harm.

Funding Sources:

Horizon Postdoctoral Fellowship, Concordia University Research Chair on Gambling Studies

Related Publications:

Murch, W.S., Scheurich, R., Monson, E., French, M., & Kairouz, S. (submitted). Distinguishing Persistent Versus Episodic Clusters of At-risk Respondents on the Problem Gambling Severity Index

Murch, W.S., Kairouz, S., & French, M. (2024). Establishing the Temporal Stability of Machine Learning Models That Detect Online Gambling-Related Harms. Computers in Human Behaviors Reports. doi: 10.1016/j.chbr.2024.100427 [link] [open materials]

Murch, W.S., Kairouz, S., Dauphinais, S., Picard, E., Costes, J.M., French, M. (2023). Using Machine Learning to Retrospectively Predict Self-Reported Gambling Problems in Quebec. Addiction. doi: 10.1111/add.16179. [link] [open materials]

Kairouz, S., Costes, J. M., Murch, W. S., Doray-Demers, P., Carrier, C., & Eroukmanoff, V. (2023). Enabling New Strategies to Prevent Problematic Online Gambling: A Machine Learning Approach for Identifying At-risk Online Gamblers in France. International Gambling Studies. doi: 10.1080/14459795.2022.2164042. [link] [open materials]

Investigating Slot Machine Immersion with Mobile Eye-Tracking Technology

I and others previously found links between immersion in slot machine use, and individual levels of problem gambling harm. However, existing theoretical models of immersion make contrasting predictions as to whether immersion results from an increasing amount of attention being paid to the slot machine (at the expense of other stimuli), or a general reduction in external attention.

In two experiments involving experienced slot machine gamblers and members of the community, I asked people to play a real slot machine while wearing a pair of eye tracking glasses. I then analyzed people’s eye movements while gambling to determine whether people who reported higher immersion in play were more — or less — interested in the outcomes of their bets.

Funding Sources:

NSERC; Centre for Gambling Research at UBC

Related Publications:

Kim, A. J., Murch, W. S., Limbrick-Oldfield, E. H., Ferrari, M. A., MacDonald, K. I., Fooken, J., Cherkasova, M. V., Spering, M., & Clark, L. (2022). Do pupillary responses during authentic slot machine use reflect arousal or screen luminance fluctuations? A proof-of-concept study. PLOS ONE, 17(7), 1-21. doi: 10.1371/journal.pone.0272070. [open access] [open data] [primer]

Murch, W. S., & Clark, L., (2021). Understanding the Slot Machine Zone. Current Addiction Reports. doi:10.1007/s40429-021-00371-x. [link] [open access]

Murch, W. S., Limbrick-Oldfield, E. H., Ferrari, M. A., MacDonald, K. I., Fooken, J., Cherkasova, M. V., Spering, M., Clark, L., (2020). Zoned In or Zoned Out? Investigating Immersion in Slot Machine Gambling using Mobile Eye Tracking. Addiction. 115, 1127-1138. doi: 10.1111/add.14899. [link] [open access] [open data] [primer]

Using Psychophysiological Technologies to Detect Biomarkers of Electronic Gaming Experiences

Past efforts to identify biological or physiological indicators of Gambling Disorder have not been successful. However, people who experience Gambling Disorder are more likely to report experiencing immersion in gambling activities, and thus any biomarker of immersion may serve as a partial predictor of problem gambling.

Across five laboratory experiments, I worked to characterize the effects of problem gambling and slot machine immersion on heart rate variability, skin conductance level, respiration rate, and cardiac impedance. In these studies, undergraduate students and members of the community were recruited to play real slot machines while wearing various physiological sensors. It is my hope that the results of these studies will help to explain the development and progression of Gambling Disorder.

Funding Sources:

NSERC; Centre for Gambling Research at UBC

Related Publications:

Murch, W. S., Ferrari, M. A., & Clark, L. (2024). Post-reinforcement pauses during slot machine gambling are moderated by immersion. Quarterly Journal of Experimental Psychology. [link] [open materials]

Murch, W. S., Ferrari, M. A., McDonald, B. M., Clark, L. (2020). Investigating Flow States and Cardiac Pre-Ejection Period during Electronic Gaming Machine Use. Frontiers in Psychology: Performance Science. doi:10.3389/fpsyg.2020.00300. [link] [open access] [open data]

Kennedy, D., Goshko, C., Murch, W. S., Limbrick-Oldfield, E. H., Dunn, B. D., Clark, L. (2019). Interoception and respiratory sinus arrhythmia in gambling disorder. Psychophysiology. 56(6), e13333. doi: 10.1111/psyp.13333. [link] [primer]

Murch, W. S., & Clark, L. (2019). Effects of bet size and multi-line play on immersion and respiratory sinus arrhythmia during electronic gaming machine use. Addictive Behaviors. 88, 67-72. doi: 10.1016/j.addbeh.2018.08.014. [link] [open access] [primer]

Murch, W. S., Chu S. W. M. & Clark, L. (2017). Measuring the slot machine zone with attentional dual tasks and respiratory sinus arrhythmia. Psychology of Addictive Behaviors, 31(3):375-384. doi: 10.1037/adb0000251. [link] [open access]