By Joseph G. Eisenhauer Misinterpretation of regression results—particularly confusion between correlation, causation, and predictive power—remains a challenge in statistics education. This paper introduces the “perfect” regression as a pedagogical device for illustrating spurious correlation and overfitting. Using extreme cases with small samples and randomly generated … [Read more...] about The ‘Perfect’ Regression: A Device for Demonstrating Spuriousness and Overfitting
Statistics Teacher
Supporting the Teaching and Learning of Statistics