Visualization of multicollinearity to show why variance inflation factors (VIFs) matter in the SE of regression coefficients.

1. Generate "triangular" data. In DAG terms, we have three variables described by the following paths:
   1. C -> Y
   2. C -> X
   3. X -> Y
   4. Importantly, X & C will be generated to have a high correlation, i.e., cor(X, C) >= 0.9
2. Make plots to show Y ~ X and Y ~ X.C (i.e., Y regressed on the residuals from X ~ C) produce same regression line, but the var(X.C) is much lower than var(X).
3. Simulate sampling from a population with the above DGM and show that the regression line from Y ~ X.C is very sensitive compared to Y ~ X.