Use the confidence intervals generated by the Dunnett method to determine whether two means are different:

·    If an interval does not contain zero, there is a statistically significant difference between the corresponding means.

·    If the interval does contain zero, the difference between the means is not statistically significant.

The simultaneous confidence level is the percentage of times that a group of confidence intervals will all include the true differences if the study were repeated multiple times.

To display the values of the confidence limits in the Session window, check Tests and confidence intervals in Stat > ANOVA > General Linear Model > Comparisons > Results.

For the salary analysis, comparisons were requested between the mean for subject 4 (the designated control level) and the means for all other levels of the subject factor. The confidence level chosen for the intervals was 95%, which corresponds to a family error rate of 0.05 (or 5%). The confidence intervals for the comparisons reveal the following:

·    The confidence interval for the difference between the means for subjects 1 and 4 contains only negative numbers, which indicates that teaching subject 4 paid significantly better than teaching subject 1.

·    The confidence interval for the difference between the means for subjects 2 and 4 contains zero, which indicates that this difference is not significant.

·    The confidence interval for the difference between the means of subjects 3 and 4 contains only positive numbers, which indicates that the mean for subject 3 is significantly greater than that for subject 4.

The 95% simultaneous confidence level indicates that we can be 95% confident that all of these confidence intervals contain the true differences.