Figure 1From: Multiple imputation of missing covariates with non-linear effects and interactions: an evaluation of statistical methods Typical datasets for normally or log-normally distributed X (each with mean 2 and variance 1), normally distributed Y with mean 2 X + X 2 or ( X - 2) 2 , and R 2 = 0.1, 0.5 or 0.8. Dotted line shows expected value of Y given X.Back to article page