EDF 7482 – Quasi-experimental design and analysis in educational research

This course examines quasi-experimental educational research designs and methods for data analysis for treatment comparisons that do not use random assignment of participants to conditions. Topics include matched designs, regression discontinuity designs, and control of selection bias using propensity scores, instrumental variables, and fixed effects.

The objectives of this course are to:

1. Familiarize students with the intricacies of selection bias in quasi-experimental designs.

2. Familiarize students with the potential outcomes framework for causal inference.

3. Provide students with the technical skills necessary to execute quasi-experimental analyses by multiple methods.

4. Familiarize students with the advantages and limitations of quasi-experimental analysis methods.

Readings will be posted on the course website. The class website is the repository for the class readings, assignments, and handouts. I will also post assignments and data sets on the website, and assignments should be submitted through the website unless different instructions are given for a specific assignment.

I will use R and RStudio software for statistical analyses. Students may choose to use SAS or STATA to complete assignments, but I will only answer programming questions with respect to R. All assignments requiring the use of statistical software should include the code used for analyses.

How to obtain and install R
1. Go to this page
2. Click on CRAN (left side panel) and select a mirror for download. 3. Select your operational system
4. Click on “Base” to download the program.

How to obtain RStudio.
1. Install R using the instructions above.
2. Go to this page
3. Follow the website’s instructions for downloading and installation.