행정학･정책학 연구에서의 이중차분 추정기법의 활용과 쟁점

초록

The purpose of this study is twofold. First, we examine the concepts and issues inherent in the difference in difference (DiD) estimator, a quasi-experimental research design widely used in empirical research in social sciences. Second, we analyze whether the DiD estimator is used appropriately in existing public administration and policy research with an aim to promote reliability and transparency of empirical research that use the estimator. We conclude the paper by emphasizing the importance of testing the “parallel trend” assumption, the identifying assumption of the DiD estimator, statistically and practically.

키워드