Arceneaux, Kevin (2010) "The Benefits of Experimental Methods for the Study of Campaign Effects." Political Communication, 27(2): 199-215. DOI: 10.1080/10584601003709407
Researchers rely heavily on observational designs to study the effects of political campaigns. In doing so, scholars often overlook serious threats to causal inference that bedevil their research designs. The strategic nature of campaigns leads to serious selection biases in voter exposure to campaign stimuli and, ultimately, generates biased estimates of campaign effects. Standard approaches to establishing causal effects in observational research, such as the collection of panel data and the inclusion of covariates, are often inadequate. In contrast, experimental approaches offer researchers a more promising way to estimate accurately the causal effects of campaigns. In this essay, I discuss the recent renaissance of experimental research in the study of campaigns and illustrate how scholars can use lab, field, and survey experiments to good effect. When experimental methods are not feasible, studies of campaign effects would benefit from paying better attention to the causal identification strategy. In closing, I discuss how the logic of experimentation can be applied to some observational settings.
Link to article here.