Abstract:
|
Estimation of causal effects in observational studies frequently requires the use of matching techniques in order to protect against bias due to a lack of balance between treated and control units. Matching methods, such as propensity score matching, were developed under the assumption of independence among units. However, modern datasets on disease prevalence, social development and even business transactions come equipped with information on a network that links the units together, rendering this assumption implausible. We provide examples of the complications that arise when information about the network is disregarded and develop a matching technique based on network attributes.
|