1. Because residuals represent the net effects of many other factors, it is rare to find a group of residuals from a simple regression that is normally distributed.
2. The nearly normal condition is critical when using prediction intervals.
3. We should exclude from the estimation of the regression equation any case for which the residual is more than 3se away from the fitted line.
4. If the Durbin-Watson statistic is near zero, we can conclude that the fitted model meets the no lurking variable condition, at least for time series data.
5. The best plot for checking the similar variances condition is the scatterplot of Y on X.
6. In regression modeling, it is more important to ft a model that meets the conditions of the SRM than to maximize the value of r2.