English (United States)
The external validity of experimental games is a growing topic of interest because researchers commonly use such games to study social preferences. Many papers have conducted within-subjects experiments, comparing in-lab game decisions to real-world decisions. However, the current accumulated evidence has yielded mixed conclusions. Some papers have found statistically significant associations between game and field behavior, while other papers have not. In this research, I compared behavior in various social preference games and self-reported measures regarding past pro-social behavior to a natural, far-removed field measure: donations to Wesleyan University. Donations were recorded as binary (whether individuals donated or not) and continuous (individuals? average yearly donations) metrics. To examine the explanatory power of social preference games on donations behavior, I ran several logistic and linear regression models. I also compared how well the games predict donations behavior relative to a baseline model. My results show that the social preference games do a poor job explaining donations behavior. On the other hand, the games have modest additional predictive power given the baseline model?s predictions. Most notably, the self-reported measures seem to perform just as well, if not better, than the social preference games in both explaining and predicting the field measure.
Xu, Sarah, "The External Validity of Experimental Social Preference Games" (2018). Honors Theses - All. 2087.
© Copyright is owned by author of this document