Workshop 2: Confidence Intervals; hypothesis testing
Read before class on Saturday, October 23, 2021
Readings
- ModernDive Chapter 4
- R4DS Chapter 12
- ModernDive Chapter 8.3-8.5
- How Obama Raised $60 Million by Running a Simple Experiment
- Coronavirus Case Counts Are Meaningless
- Bertrand M, Mullainathan S. (2004). Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination. The American Economic Review 94:4 (991-1013).
Questions to reflect on
- In the Bertrand and Mullainathan (2004) experiment, the authors randomly assigned first names, and thus race and gender, to fictitious candidate resumes submitted to online job ads. Since other candidate attributes and overall candidate quality were assigned independent of name, the authors wanted to measure differences in callback rates for interviews.
- Would you expect to see any statistical differences in callback rates between races? The authors state that Job applicants with white names needed to send about 10 resumes to get one callback; those with African-American names needed to send around 15 resumes to get one callback.
Session Files
- You can download all session files (data, code, etc.) by pulling from course Github repo.
Alternatively, please install package
usethis. Once you have it, you can download, unzip, and open everything within an RStudio project by typing the following in the RStudio console
usethis::use_course("https://github.com/kostis-christodoulou/usi_EMBA_analytics/raw/master/session4-workshop2.zip")