B1180
Title: A representative sampling method for peer encouragement designs in network experiments
Authors: Yanyan Li - University of Southern California (United States)
Qing Liu - University of Wisconsin-Madison (United States) [presenting]
Sha Yang - University of Southern California (United States)
Abstract: Targeted marketing interventions are prevalent on social networks. Firms are increasingly interested in conducting network experiments through peer encouragement designs to causally quantify the potentially heterogeneous direct effect of a marketing program on focal individuals (egos) and the indirect effect on those connected to the focal ones (alters). A widely adopted practice to obtain clean estimates of the direct and indirect treatment effects in peer encouragement designs is to draw random samples from the population network and then exclude contaminated egos and alters from the inference. However, the approach may lead to underrepresentation and undersupply of the resulting treatment/control samples. A Bayesian representative sampling algorithm is proposed to improve the peer-encouraged designs and the related causal inference. Through simulations, it is shown that, compared with those obtained from the post hoc excluding approach, samples constructed based on the proposed method allow researchers to more precisely estimate the average treatment effects and the heterogeneity in individual treatment responses and predict the treatment effects out of h sample. Moreover, the proposed method is computationally efficient and can be conveniently adapted and incorporated into many applications for evaluating social influences.
