`## Validity`

The residuals should be independent, normal random variables with a mean of zero and constant variance.


`### Distribution of Residuals`

For the inferential analyses to be valid, the residuals should be normally distributed. Violation of normality does not bias the estimates, but it does render the inferential tests invalid.

`plot.resids_dist`



`### Fitted Values vs Residuals`

The patterning of the residuals should represent random variation, free of any pattern or structure. They should satisfy the _homoscedasticity_ assumption, randomly scattered about 0, with approximately the same level of variability across the range of the fitted values within a horizontal band around the zero-line. Otherwise they demonstrate _heteroskedasticity_.

`plot.fitted_resids`