Ken Blake, Ph.D.

Measuring linear association between two variables: Bivariate regression.

Lots of things are associated with obesity. Heavy television viewing is one of them, mainly because watching television tends to be a sedentary activity, and the more sedentary one is, the more likely one is to gain weight.

This example uses made-up data to show how regression can quantify the strength of the association between hours per week spent watching television and percentage of body fat and check on whether the association is strong enough to suggest it is nonrandom. There's also some important information about how to avoid getting fooled by unusual cases called "outliers."


Ready to give it a try? Here's the dataset.