simplest way to compare means from independent samples is usingconfidence intervals, as shown in the previous section. In comparing the confidenceintervals for two means, you can draw the following conclusions:don’t overlap, you can safely assume the twomeans are significantly different from each other (at the confidence levelyou chose).overlap slightly, the two means might still besignificantly different. Run attest to determine if they are different.overlap widely, the two means are not significantlydifferent.Let’s consider the data inFigure 2.3to illustrate running attest for independentsamples. This shows the ratings of ease of use on a 1 to 5 scale fortwo different designs as rated by two different groups of participants (who wereassigned randomly to one group or the other). We’ve calculated the means and-vals overlap slightly: Design 1’s interval goes up to 3.8, whereas Design 2’s goesdown to 3.5. This is a case where you should run attest to determine if the twomeans are significantly different.