When you select the Multiple Comparisons option, you can choose the initial comparison to be with a control group. If you select Comparisons with Control - Dunnett’s, a window opens, asking you to specify a control group. If you selected Least Squares Means Estimates, the list consists of all levels of the effect you that you selected. If you selected User-Defined Estimates, the list consists of the combinations of effect levels that you specified.
Tip: You can add a Control Level column property to the factor column to avoid specifying the control group each time you select Comparisons with Control - Dunnett’s. See “Control Level” in Using JMP.
After you choose a control group and click OK, the Comparisons with Control report appears in your Fit Least Squares report. This option compares the means for the specified settings to the control group mean. It displays a table showing confidence intervals for differences from the control group and a chart showing decision limits. Dunnett’s method is used to make the comparisons. Dunnett’s method is a multiple comparison procedure that controls the error rate over all comparisons (Hsu 1996; Westfall et al. 2011).
When exact calculation of p-values and confidence intervals is not possible, Hsu’s factor analytical approximation is used (Hsu 1992). Note that computing exact critical values and p-values for unbalanced designs requires complex integration and can be computationally intensive. When calculations for such a quantile fail, the Sidak quantile is computed.
In addition to the list of effects that do not vary for the specified estimates, at the top of the Comparisons with Control report you also find:
Quantile
The critical value for Dunnett’s test.
Adjusted DF
The degrees of freedom used in constructing the confidence intervals.
Control
The setting that defines the control group. This is a single level if you have selected a single effect; it is a combination of levels if you specified a user-defined combination of more than one effect.
Adjustment
The method used to obtain the critical value:
Dunnett
Provides exact critical values and p-values. Used whenever possible, in particular, when the estimates are uncorrelated.
Dunnett-Hsu
Provides approximate critical values and p-values based on Hsu’s factor analytical approximation (Hsu 1992). Used when exact values cannot be obtained.
Sidak
Used when both Dunnett and Dunnett-Hsu fail.
For technical details, see the GLM Procedure chapter in SAS Institute Inc. (2024b).
Three options are available from the Comparisons with Control report menu:
For each comparison of a group mean to the control mean, this report provides the following details:
• The levels being compared
• Difference - the estimated difference
• Std Error - the standard error of the difference
• Lower and Upper limits for the confidence interval
• t Ratio - the ratio of the Difference and Std Error columns
This decision chart plots a point at the mean for each group being compared to the control group. A horizontal line shows the mean for the control group. Upper and lower decision limits are plotted. When a point falls outside these limits, it corresponds to a group whose mean differs from the control group mean based on Dunnett’s test at the specified significance level. That level is shown beneath the chart.
The Comparisons with Control Decision Chart report menu has these options:
Show Summary Report
Produces a table showing the estimate, decision limits, and the limit exceeded for each group
Display Options
Provides several options for controlling the display of the chart.
Adds a column that contains p-values (Prob>|t|) to the Comparisons with Control report. Note that computing exact critical values and p-values for unbalanced designs requires complex integration and can be computationally challenging. When calculations for such a quantile fail, the Sidak quantile is computed but p-values are not available.