QSS DUR harnesses machine learning to analyze complex experimental designs
Assistant Professor of Government Sean Westwood and coauthors recently released a new statistical method that harnesses machine learning to analyze complex experimental designs. Their method was published in Political Analysis, the top methodology journal in political science. With his colleagues, Westwood, the Director of Undergraduate Research in the Program in Quantitative Social Science, showed that it is possible to utilize a super-learner to estimate the effects of heterogenous treatments on respondents' attitudes. Their method also allows for the analysis of heterogeneous responses to treatments.