The summary plot shows the multiple comparison intervals for the standard deviations. The multiple comparison intervals are not the same as the Bonferroni confidence intervals (CIs). Use each type of interval as follows:

·    Use the Bonferroni simultaneous confidence intervals to jointly estimate the standard deviation of each population.

·    Use the multiple comparison intervals to determine which standard deviations are significantly different from each other. If two intervals do not overlap, then the corresponding standard deviations (and variances) are significantly different.

If the p-value for the multiple comparisons test is less than your chosen significance level, then at least one standard deviation is significantly different from one other standard deviation.

For the driving data, the p-value for the multiple comparisons test is much larger than the significance level of 0.05. There are no significant differences between groups, and all of the comparison intervals overlap.

However, there are only 4 observations for each group. With such small sample sizes, the tests do not have much power to detect a difference if one does exist. You should consider collecting more data to increase the power of the test.

In general, you should base your conclusions on the multiple comparisons test and corresponding multiple comparison intervals, unless you have small samples from very skewed, or heavy-tailed distributions. (See Tests of equal variance for further discussion.)