|
Abstract:
|
Classical approaches to causal inference largely rely on the assumption of "no interference", according to which the outcome of an individual does not depend on the treatment assigned to others. In many applications, however, such as evaluating the effectiveness of healthcare interventions that leverage social structure, or assessing the impact of product innovations on social media platforms, assuming lack of interference is untenable. In fact, the effect of interference itself is often an inferential target of interest, rather than a nuisance. In this lecture, we will formalize technical issues that arise in estimating causal effects when interference can be attributed to a network among the units of analysis, within the potential outcomes framework. We will then introduce and discuss several strategies for experimental design in this context.
|
