Exact Confidence Intervals in the Presence of Interference

• Activity Number: 33
• Type: Contributed
• Date/Time: Sunday, August 9, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Joseph Rigdon* and Michael Hudgens
• Companies: Stanford University and The University of North Carolina at Chapel Hill
• Keywords: Causal Inference ; Exact Confidence Interval ; Interference ; Randomization Inference

Abstract

Interference is present when the outcome of one individual is affected by the treatment of other individuals. Partial interference is a special case of interference where individuals can be partitioned into groups such that there is no interference between individuals in different groups. Two-stage randomized experiments have been proposed for drawing inference about treatment effects when partial interference is assumed. For a two-stage randomized experiment assuming stratified interference, in this talk methods are developed for constructing exact confidence intervals for the direct, indirect, total, and overall effect of a treatment on a binary outcome. The methods are nonparametric and require no assumptions about random sampling from a larger population. The confidence intervals are exact in the sense that the probability of containing the true treatment effects is at least the nominal level. The new exact confidence intervals are compared via simulation with previously proposed exact and asymptotic confidence intervals. While the asymptotic intervals do not maintain nominal coverage for certain simulation setups, the new exact confidence intervals maintain nominal cover