Authors: Elizabeth Diemer, Joy Shi and Sonja Swanson
The Data Challenge aims to highlight some of the methodological challenges of inferring causal effects from real-world data using Mendelian Randomization (MR) and provide a concrete anchor for discussions of these complexities among conference attendees. In particular, we encourage participants to consider issues of selecting potential instruments, clearly defining causal estimands, and timing. Here, you'll find the following files needed to complete the challenge:
- Instructions and prompts for the data challenge
- Two datasets: one corresponding to Part 2 of the data challenge and one corresponding to Part 3 of the data challenge
- Data dictionary describing the variables in each of the datasets
After the challenge is over, descriptions of the data generating model and the code used to generate the data will be provided.