In this week’s episode, Greg and Patrick talk about the terrifying, the feared, the dreaded … Multicollinearity. Blamed for a multitude of general linear model problems, they dare to ask the question: “But should it be?” Along the way they also mention: having your stump ground out, fall guys, Keyser Soze, croissants and breadsticks, baguettes in space space space, mostly dead, the Cliffs of Moher, enablers, dangling on a wing and a prayer, nanotech R-squareds, opening a suitcase, reinventing factor analysis, and whiny a** babies.
Related Episodes
S4E01: Ordinary Least Squares: Back Where It All Began
S3E09: Semi-Partially Clarifying Measures of Association in Regression
S2E18: Regression: Like That Old High School Friend You’ve Outgrown
S2E14: Control (Variable) Issues
Recommended Readings
Alin, A. (2010). Multicollinearity. Wiley Interdisciplinary Reviews: Computational Statistics, 2, 370-374.
Daoud, J. I. (2017, December). Multicollinearity and regression analysis. In Journal of Physics: Conference Series (Vol. 949, No. 1, p. 012009). IOP Publishing.
Graham, M. H. (2003). Confronting multicollinearity in ecological multiple regression. Ecology, 84, 2809-2815.
Mansfield, E. R., & Helms, B. P. (1982). Detecting multicollinearity. The American Statistician, 36, 158-160.
Shrestha, N. (2020). Detecting multicollinearity in regression analysis. American Journal of Applied Mathematics and Statistics, 8, 39-42.