In this chapter, we examined how to develop functions that perform simple regression analyses, and how to multiply regression in R using a real life example. We have examined the importance of significance tests for regression, and have briefly discussed robust regression and bootstrapping. Note that, when data about the predictors and the criterion are collected simultaneously, causation cannot be established. In order to ascertain causation, data must be collected longitudinally—that is, the predictors before the criterion.

In the next chapter, we will examine the classification of observations using the Naïve Bayes and k-Nearest Neighbor algorithms.