This is from a blog post by Professor Lucas Davis at his blog. In addition to the subject, that’s an interesting way of presenting a change over time I’ll need to think about: It seems the model could be used in other, more comprehensive ways. Note it’s really a matched pairs test, where each state is a candidate and its electricity use in 2010 is match with that in 2015. Even though the amount of electricity used by any individual state over time is a dependent quantity, electricity use of one state is more or less independent of that in another state. They might be dependent if, say, the United States economy crashed, or if it underwent a sudden boom.