Episode 3｜Association, Inference, and Causal Thinking in Simple Linear Regression

This episode builds on simple linear regression by focusing on statistical inference—how we move from a fitted line to meaningful conclusions. We review the intuition behind least squares and explain why switching the roles of the two variables can lead to different fitted lines.We then discuss how to interpret regression results in practice, including point estimates, uncertainty, and the difference between statements about an average outcome versus predictions for an individual. Using simple examples and R-based illustrations, the episode highlights how confidence intervals and prediction intervals answer different questions.Finally, we return to a key warning in applied data analysis: association is not causation. Through classic real-world examples (such as ice cream sales and shark attacks), we explain how hidden variables can create misleading relationships—and why a strong regression fit does not automatically justify a causal claim.