Business Knowledge Series course

Presented by Howard S. Friedman, Ph.D., Professor, Columbia University, and Partner, DataMed Solutions LLC

This course introduces some methods commonly used in program evaluation and real-world effectiveness studies, including two-stage modeling, interrupted time-series, regression discontinuity, and propensity score matching. These methods help address questions such as: Which medicine is more effective in the real world? Did an advertising program have an impact on sales? More generally, are the changes in outcomes causally related to the program being run?

Learn how to

• identify alternative techniques to propensity score-based ones, including those that require different two-stage setups (Heckman) and time-series techniques
• apply quasi-experimental analysis methods to real-world data for the following techniques: propensity score matching, Heckman's two-stage model, interrupted time series, and regression discontinuity.

Who should attend

Data analysts, statisticians, and economists in the fields of finance, telecommunications, pharmaceuticals, and retail and in the public sector, who have an understanding of basic statistics and SAS programming. Those who work in areas of economics, program evaluation, and real-world effectiveness studies will find this course highly relevant.

Formats available	Standard duration
e-Learning:	14 hours/180 day license	System Requirements

Before attending this course, you should have completed the following courses or have the equivalent working experience:

• SAS Programming 1: Essentials
• Statistics 1: Introduction to ANOVA, Regression, and Logistic Regression.

Prior study of multiple linear regression modeling is required.

This course addresses Base SAS, SAS/ETS, SAS/STAT software.

Heckman's Two-Stage Model

• Heckman's two-stage modeling
• basic concepts of randomized experiments
• sample selection versus treatment selection
• introduce Heckman's two-stage model
• applying Heckman's two-stage model to a data set on wages

Propensity Score Matching

• introduction
• real-world non-random treatment assignment
• propensity score matching basics
• interpreting the propensity matched results

Interrupted Time Series

• time series introduction
• basic concepts of time series analysis
• types of interrupted time series models
• interrupted time series example
• other interrupted time series analysis

Regression Discontinuity

• regression discontinuity introduction
• real-world examples of regression discontinuity
• motivation behind regression discontinuity
• detecting a regression discontinuity
• regression discontinuity example

Participant Examples and Course Summary (included only in classroom version)

• key situations when methods might be applied
• discuss key situations when methods might be applied
• discuss examples of real-world quasi-experimental designs