Causal Analysis and Inference

Causal Analysis and Inference (PAF 9272) is the second course in the research methods sequence. Upon completion of Data Collection and Description (PAF 9270), students can choose between this course and Data Analysis for PublicService (PAF 9271). PAF 9272 is meant for those interested in becoming analysts, researchers, or making quantitative data analysis an important element in their careers. It teaches students how to critically evaluate existing causal analyses of both qualitative and quantitative data and how to conduct statistical analyses to answer causal questions for domestic and international policy and practice. Compared to PAF 9271, PAF 9272 places greater emphasis on observational and experimental data from representative surveys and requires students to write programs (coding) to carry out statistical analyses using advanced statistical software such as Stata or R. The course provides a hands-on introduction to understanding causal evidence, covering logic models and mechanisms, case-oriented vs. variable-oriented approaches, correlation vs. causation, observational vs. (quasi) experimental data, treatment effect, confounding and omitted variable bias, complex survey sampling, generalizability, standard error, confidence interval estimation, hypothesis testing, statistical and practical significance, power analysis, multiple regression, and difference-in-differences estimation. Course sections will use applications tailored towards students’ interests and concentrations (e.g., sections more populated withMIA students will have a greater international focus). (Students who took PAF 9170 or PAF 9172 cannot get credit for this course. They, and all other students looking for an advanced causal methods and statistics course, should considertaking PAF 9177.) Open to Austin W. Marxe School of Public and International Affairs MPA and MIA students; others with Marxe School permission.

035520