Linear multiple regression, aka regression, is both the over-stressed workhorse of multivariate analysis and the primary vehicle by which the basic logic of 21st century multivariate research can be conveyed. Our aim in this class is therefore two-fold.
First, we seek to understand regression analysis as a potentially useful tool. This will necessitate attending to the statistical assumptions of the model, diagnostic tests of those assumptions, possibly corrective interventions when assumptions are violated, and a critical appraisal of the statistical model in light of the various aims a researcher may have.
Second, we seek to build our understanding of regression in a way that opens the door to understanding any of the other methods that have been developed in the last quarter century, methods that increasingly dominate the statistical analysis of social phenomena. This effort will not take us into study of those approaches, but it will take us to the brink of that endeavor, providing sufficient grounding for the future study of those methods, or the critical evaluation of work based on those methods.