Inside Outside: The interaction between immersive tendencies and sense of presence in changing risk-taking behavior via serious games

Carlos Godoy, John Christensen, Lynn Miller, Jennifer Talevich, Paul Robert Appleby, Charisse Corsbie-Massay and Stephen Read

Extended Abstract

Serious Games have great potential to optimize reduction in individual risky decision-making and behavior (Miller et al., in press). It's critical to know, however, what works when for whom. One key to their success may be that individuals may treat the virtual experience "as if" it is real life (Godoy, 2007). Individuals' experience of feeling immersed in the ongoing interaction may facilitate this. Such a propensity may be due to both measurable chronic individual differences (e.g., via the Immersive Tendencies Questionnaire (ITQ); Witmer & Singer, 1994) as well as induced changes due to the specific visual and narrative characteristics of the interactive media itself (e.g., as measured by the Presence Questionnaire Revised (PQR); Witmer & Singer, 1994). We predicted that chronic immersive tendencies and/or an induced "sense of presence", individually or in combination, would predict reduction in risky sexual decision-making in the 3-months following exposure to an HIV prevention serious game intervention.

107 men who have sex with men (MSM) participated in a longitudinal study assessing an interactive video (IAV) behavioral intervention designed to reduce sexual risk-taking. The intervention placed MSM in a virtual environment that realistically simulated a sexual encounter while attempting to train advantageous decision-making via traditional cognitive as well as emotion-based approaches (Miller et al, in press). MSM who had engaged in risky unprotected anal intercourse (UAI) with a non-primary male partner within the past 90 days were recruited at venues or via internet. Measures of prior UAI were assessed at baseline. Participants then received the IAV intervention, and later completed the PQR. At three month follow-up, MSM reported their UAI behavior and completed the ITQ.

A hierarchical regression was conducted to predict change in UAI from presence and immersive tendencies. Neither presence nor immersive tendencies significantly predicted UAI change in the first step, F(2, 107) = 2.59, p > .05. However, the interaction between immersive tendencies and presence was a significant predictor in the second, F(3, 107) = 5.11, p < .01. The model including this interaction term predicted 13% of the variance in UAI risk-taking change. When plotted against the mean for regression line at 0, we found no significant difference in UAI change between low immersives who experienced a high sense of presence and those who experienced low presence. The intervention worked well, however, for highly immersive individuals who also reported high levels of presence within the virtual environment.

The current findings suggest that, although presence may be a less important consideration for those low in immersive tendencies, those who have a natural propensity to become highly immersed are best served by serious games that maximize feelings of presence.