Reducing HIV Risk Nationally for Young Men Who Have Sex with Men: Developing SOLVE-It (Socially Optimized Learning in Virtual Environments Using Intelligent/Gaming Technologies)

Theoretical Background and research questions/hypothesis:  Traditional HIV prevention interventions for MSM have reduced Unprotected Anal Intercourse (UAI) but these more conscious, deliberative, and cognitive approaches do not address a more automatic, affect-based route to decision-making. Experience with risk cues is needed to produce such automatic, affect-based risk reduction: But, real-life experience could be catastrophic. SOLVE integrates traditional cognitive approaches (e.g., social-cognitive interventions modeling cognitive and behavioral skills), while addressing MSM's affect- based and reactive risky decision-making processes, by giving them experience with risk cues in a safe, virtual environment. SOLVE has been found to reduce UAI compared to a wait-list control for three ethnic populations (Black/African-American, Latino/Hispanic, White/Caucasian) of younger (18-24 year old) MSM. However, interactive video technology (SOLVE-IAV) is limited. SOLVE-IT, an NIMH funded serious game with intelligent agents, provides a richer set of challenging, more personalized, interpersonal risk situations (with a future virtual self who guides the character's decision making). Game development requires a series of pilot studies addressing research questions. RQ: How should we frame messages when users have made a safer or riskier choice? RQ: Does affect reporting during the game interrupt user's sense of presence or influence intentions to use condoms?

Methods:  Designed for younger MSM from our three high risk groups, users can modify their character's appearance in the game to better match their own and choose their own responses to their romantic partners at a bar, at a house party and in their partner's apartment. The game is being developed over a 32-month period with a team of social scientists and artificial intelligence, gaming, and animation experts. Formative pilot studies informed a series of decisions regarding how to measure affect, and what interpersonal and other challenges and which interventions to include in the game. SOLVE-IT will be tested in a 6-month longitudinal randomized controlled trial nationally over the web, where UAI change is the primary dependent variable.

Results:  Pilot study analyses revealed that (1) affect measured intermittently during the game was not disruptive to the player and could provide more diagnostic cues to subsequent intention to use condoms and (2) the effects of loss or gain frame messages given after a risky or safer choice differ substantially from what conventional theory might predict. The SOLVE-IT development process, pilot results, and lessons learned, are described. Clips and videos that provide a feel for a shortened part of the game will be played.

Conclusions:  Researchers cannot just rely on past theory and research designed for more passive media interventions in designing message interventions or anticipating user reactions within interactive environments. Formative research is critical in designing games.

Implications for research and/or practice:  Implications of pilot studies and designing interactive intelligent technologies/games for changing risky behavior are discussed.