11. False. When the variance of the errors changes, prediction intervals are likely to be too short in some cases and too long in others.
17. False. If omitted variables contribute small, independent deviations, their net effect tends to have bell−shaped variation.
19. False. The decision to exclude data should account for substantive relevance. An unusual case might be the most important data.
21. False. It is easier to see changes in the variation in the plot of residuals on x.