This view is a selection tool for 2-level factorial designs (we recommend using the Half-Normal Plot of Effects.)
Select the terms to go into the ANOVA model by clicking on the effects squares. Start at the extremes (left and right side) working towards the center of the plot. Select effects that fall off the red line.
The Shapiro-Wilk test is testing the unselected terms to determine if they follow a normal distribution. The Shapiro-Wilk p-value should be greater than 0.10 after the important effects have been selected. This means the unselected effects are approximately normally distributed. If the Shapiro-Wilk p-value is greater than 0.10 with no terms selected, then it may be possible that there are no statistically significant effects.
Ctrl + right-click will show the x-axis effect values that are currently displayed.
Shift + right-click will show the y-axis percentiles that are currently displayed.
For either of the above, the values shown are copied to the clipboard to be pasted into another program.