Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
This lecture on observational studies by the instructor covers the difference between experimental and observational studies, highlighting the pitfalls of the latter. It provides tools to avoid these pitfalls, allowing for valid conclusions from 'found data'. The lecture aims to motivate students to read key books in the field. It discusses randomized controlled experiments, the limits of randomization, and the importance of observational studies. The concept of matching in observational studies is explained, along with the use of propensity scores to address biases. The sensitivity analysis model is introduced to quantify the impact of unobserved covariates. Examples related to smoking and lung cancer are used to illustrate the concepts discussed.