Statistical Workflow and the Fractal Nature of Scientific Revolutions

Explore the complexities of statistical workflow and scientific revolutions in this thought-provoking lecture by Andrew Gelman from Columbia University. Delve into the challenges of automated inference and best practices in statistical problem-solving, moving beyond single model inference. Examine how these issues apply to human research teams designing and analyzing quantitative data across various application areas. Gain insights into how an A.I. might approach statistics, considering model building, checking, and revision processes. Discover the potential parallels between automated inference and human-led statistical practices, and contemplate the fractal nature of scientific revolutions in the context of modern data analysis.