Overview
Many courses teach you how to analyze data, but few courses teach you how to ensure that you have the right data to analyze in the
first place. And few courses teach you how to interpret results in a way that makes appropriately qualified claims about generality,
causality, and the potential effects of future interventions. This course emphasizes fundamental issues including sampling, causality,
generalizability, significance, and effect size. We examine research designs that are meant to enhance causal inference, including
randomized experiments, regression discontinuity, propensity score matching, and “crossover” longitudinal designs with fixed effects
and instrumental variables.
Prerequisites
We assume acquaintance with basic statistical inference including differences between means (e.g., t tests), differences between
proportions (e.g., chi-square tests), and linear regression. Previous familiarity with Stata will be helpful but is not required.