Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore the concept of geocentric models in statistical thinking through this lecture that examines how seemingly reasonable but fundamentally flawed models can still produce accurate predictions while missing the underlying causal structure. Learn about the historical parallels between Ptolemaic astronomy and modern statistical modeling, understanding how complex epicycles in ancient geocentric models mirror the overfitting and complexity issues in contemporary data analysis. Discover why models that fit data well may still be wrong about causation, and develop critical thinking skills for evaluating statistical models beyond their predictive accuracy. Examine the philosophical and practical implications of model selection, the difference between description and explanation in statistical modeling, and how to avoid the trap of mistaking correlation for causation. Gain insights into the importance of causal thinking in statistics and learn to distinguish between models that merely describe patterns in data versus those that reveal true underlying mechanisms.
