In this repo I demonstrate the paradox of propensity score matching using the Teleco Churn Dataset.
- PSM (Rubin & Rosenbaum)
- the paradox (King & Nielsen)
- Clone the repo
- Download the dataset (I got it from kagglehub) - name it email_analytics.csv
- Run the imports