Correlation & Regression

This lesson is part of Premium

Correlation coefficient r (−1 to +1): strength & direction of linear association; Pearson (parametric) vs Spearman (rank/non-param). r² = variance explained. Correlation ≠ causation. Regression: predicts outcome from predictor(s); linear vs logistic (binary outcome→OR) vs Cox (survival→HR). Confounding adjustment via multivariable regression. Diagram: scatter plots at various r. Exam tip: strong correlation can be wholly confounded.

Premium unlocks all 40 lessons, every calculator and the timed mock.