A Gentle Introduction to Causal Inference

Explore the fascinating world of causal inference in this 53-minute talk from the EuroPython 2021 conference. Delve into the concept that correlation does not always imply causation and discover ingenious techniques for unveiling causal relationships within observational data without resorting to expensive random control trials. Learn about basic causal inference concepts through accessible visualizations and understand how adding this tool to your statistical arsenal can lead to better experiments and more insightful data analysis. Examine Simpson's Paradox to highlight the importance of using graphs to model data and manage confounding factors. Aimed at anyone involved in data-driven decision-making, this talk emphasizes the significance of understanding the story behind the data. Gain valuable insights into going beyond simple correlation calculations, extracting more meaningful information from your datasets, and avoiding common misinterpretation pitfalls. By the end of this gentle introduction, you'll be inspired to further explore causal inference and its potential to revolutionize your approach to data analysis.