Overview
I was given the opportunity to design and pilot a new Stata Lab course for my department, which is now a degree requirement for the Tulane Economics B.A. and B.S. programs. In Stata Lab, students learn basic data analysis and how to implement statistical tests in Stata, a popular data analysis software for social scientists. The course follows the topics of the undergraduate econometrics course (taken concurrently), teaching students how to perform the statistical inference they learn theoretically with actual data. I was given the freedom to create the structure of the course as well as the lectures, topics-covered (given that these topics included the statistical tests learned in econometrics), assignments, and exams.
I decided that Stata Lab should emphasize learning by trial and error and clarifying the connection between econometric theory and Stata code and ouput. Each week, the first class was a lecture and the second was an in-class activity, where students worked in groups to answer questions related to that week's lecture. As this was most of the students first experience coding in any respect, it was essential for students to build experience and comfort using a data analysis software. In-class activites were invaluable for this end, allowing students to try, fail, and succeed in Stata in a comfortable environment with the instructor nearby for questions and support. To encourage exploration, in-class activities were graded on the basis of effort rather than correctness, with answer keys provided afterwards to explain concepts, how to set up Stata code, and interpret output. Learning to code is a daunting process, and this grading structue allowed students the freedom to attempt to use Stata with as little pressure as possible.
As this course prepares students to use data analysis to answer economic questions, I stressed the importance of validly interpreting Stata output from statistical tests. Focusing on intuition rather than pure math, I highlighted the assumptions underpinning inference and how to learn about the real world using data analysis. This led me to give overviews of econometric concepts that focused on setup and interpretation, which students found helpful not only for Stata Lab, but for their econometrics course as well. By the end of the course, students have the tools to take data they find in the real world, import it into Stata, and use their data analysis skills to answer basic questions.
Examples of Class Slides and Assignment .do Files
In-Class Activity and Answer Key: Estimating the Effects of Passenger Sex and Class on the Probability of Survival on the Titanic