This course available at no cost -

AERA Institute on Statistical Analysis for Education Policy: Causal Inference

Causal inference has become a central issue in education research. This includes inferring causality from the design of true randomized experiments, as well as how such inference can be approached in quasi-experimental, non-randomized studies. The focus of the Institute is on these issues and the methodologies available to support causal inferences. The Institute covers topics such as the design of randomized experiments and the difficulties of implementing them in educational settings as well as several approaches and methodologies for estimating causal inferences in situations where randomized studies are impossible or too costly. Such methodologies include propensity scores, regression discontinuity, instrumental variables, path analysis, and structural equation models, as well as related sensitivity methods. During the Institute examples are provided implementing selected methods with data.

Faculty:

• Ken Frank, Michigan State University
• Richard Houang, Michigan State University
• Kim Maier, Michigan State University
• William Schmidt, Michigan State University

Course sessions:

• Introduction to Data Set: National Board Certified Teachers
  Presented by Ken Frank, Michigan State University
  (43 Minutes)
• Randomized Experiments
  Presented by Kim Maier, Michigan State University
  (62 Minutes)
• Causal Modeling Outside the Randomized Experiment
  Presented by William Schmidt, Michigan State University
  (80 Minutes)
• Regression Modeling Approach to Causal Inference
  Presented by Ken Frank, Michigan State University
  (83 Minutes)
• Propensity Score Modeling Approach to Causal Inference
  Presented by William Schmidt, Michigan State University
  (85 Minutes)
• Example and Data Analysis Related to Propensity Score Modeling
  Presented by Richard Houang, Michigan State University
  (47 Minutes)
• Sensitivity Analysis: Quantifying the Discourse about Causal Inference
  Presented by Ken Frank, Michigan State University
  (82 Minutes)
• Discontinuity Modeling and Instrumental Variable Modeling Approach
  Presented by Kim Maier, Michigan State University
  (58 Minutes)
• Path Modeling and Structural Equation Modeling Approach to Causal Inference
  Presented by William Schmidt, Michigan State University
  (75 Minutes)