Exercise

Likelihood-ratio test

This exercise features a new concept that I haven't covered yet, but I think you are ready for it. Likelihood ratio tests are used to compare two models. They tell us whether one model fits the data better than another model, and you can perform this using the lrtest() command, which takes as input two different model objects. We can use it to test whether choc_m2, which has a linear price coefficient, fits the data better than choc_m3, where we treated price as a factor. Those models are already loaded.

Instructions

100 XP

Perform a likelihood-ratio test comparing the models choc_m2 and choc_m3.