37  Traditional Tests

As you have read inprevious chapters, we are super enthusiastic about using randomization approaches to inference. Nevertheless, there are some situations (such as when reading a scientific journal) where it may be useful to know about traditional “frequentist” tests and estimates.

This chapter gives you the briefest possible introduction to testimate, a CODAP plugin that performs these traditional tests. You’ll find it in the Plugins menu under Statistcial Inference.

When you want to know more, see this guide to the plugin.

The live example below is all set up for you to perform a test. We see NHANES data on 20 individuals, including BMI and Sex. Is the mean BMI the same for the two genders? The graph suggests that females have a higher mean value.

1. Drag the outcome variable BMI from the table into the test, where is says drop attribute here. A test appears, testing whether BMI could be zero. Answer: no way. We move on.
2. Drag the predictor, Sex, to the right-hand spot in the test. The test does not change, but a menu appears, because more tests are available now.
3. Use the menu to change the test to difference of means. Now we see a P-value of about 0.04.
4. Play with the various controls! Do a one-way ANOVA! Change the data and see the test update!