On Teaching Correlation and Regression.

The purpose of this article is to present an example with numbers sufficiently small that the mathematics can be reasonably handled by the students. This example is designed to adequately demonstrate, in a convincing way, a large number of important concepts in correlation and regression analysis. Social statistics courses vie for the number one position as the most hated course in the social science curriculum. Perhaps the major parameter determining the repugnance of statistics, other than the inherent tediousness and lack of "pop-interest" of the material, is that students are afraid of numbers. When the student becomes flustered at the most simple mathematical calculation, such examples are hard to come by. If an example is too simple, the students can follow the mathematical procedures, but the example does not adequately reflect statistical ideas. If an example adequately reflects the statistical notions, the students get lost in the mathematical procedures. When the student completes the statistics course, he must be convinced of a number of principles of correlation and regression analysis.