A growing number of scholars and data
scientists are conducting randomized experiments to analyze causal
relationships in network settings where units influence one
another. A dominant methodology for analyzing these network
experiments has been design-based, leveraging randomization of
treatment assignment as the basis for inference. In this paper, we
generalize this design-based approach so that it can be applied to
more complex experiments with a variety of causal estimands with
different target populations. An important special case of such
generalized network experiments is a bipartite network experiment,
in which the treatment assignment is randomized among one set of
units and the outcome is measured for a separate set of units. We
propose a broad class of causal estimands based on stochastic
intervention for generalized network experiments. Using a
design-based approach, we show how to estimate the proposed causal
quantities without bias, and develop conservative variance
estimators. We apply our methodology to a randomized experiment in
education where a group of selected students in middle schools are
eligible for the anti-conflict promotion program, and the program
participation is randomized within this group. In particular, our
analysis estimates the causal effects of treating each student or
his/her close friends, for different target populations in the
network. We find that while the treatment improves the overall
awareness against conflict among students, it does not significantly
reduce the total number of conflicts. |