A GEOMETRIC APPROACH TO COMPARE VARIABLES IN A REGRESSION-MODEL

Geometry is a very useful tool for illustrating regression analysis. D espite its merits the geometric approach is seldom used. One reason fo r this might be that there are very few applications at an elementary level. This article gives a brief introduction to the geometric approa ch in regression analysis, and then geometry is used to shed some ligh t on the problem of comparing the ''importance'' of the independent va riables in a multiple regression model. Even though no final answer of how to assess variable importance is given, it is still useful to ill ustrate the different measures geometrically to gain a better understa nding of their properties.