Reconciling statistical power, stylization, and tailoring in framed field experiments and serious games

The related research areas of framed field experiments and participatory modelling stake out contrasting poles in the approach of informing models via behavioural experiment. Along the spectrum between them, researchers make trade-offs in simplicity (for improved statistical power) and specificity (for better representation and participant engagement). Where researchers land along this spectrum shapes the generalizability and external validity of their findings. Additionally, wherever researchers land along this spectrum, they have produced a ‘game stylization’ that differs from the perceived reality of their participants, and may still differ from the ‘model stylization’ through which processes are implemented in their model. I present a framework for building spatial games in Netlogo, demonstrate the simplicity-specificity problem via several different game applications, and discuss approaches (such as flexible game module toolkits and community planning sessions, through which participants contribute to designing the game structure and rules) that may help to reconcile some of the conflicting challenges in developing games that connect to participants, to models, and to statistical analysis alike.