CONTINUUM REGRESSION AND RIDGE-REGRESSION

We demonstrate the close relationship between first-factor continuum r egression and standard ridge regression. The difference is that contin uum regression inserts a scalar compensation factor for that part of t he shrinkage in ridge regression that has no connection with tendencie s towards collinearity. We interpret this to mean that first-factor co ntinuum regression is preferable in principle to ridge regression if w e want protection against near collinearity but do not admit shrinkage as a general principle. Furthermore, our experience indicates that wi th first-factor continuum regression we can obtain predictors that are at least as mean-squared error efficient as with ridge regression but with less sensitivity to the choice of ridge constant. The scalar com pensation factor is easily calculated by just an additional simple lin ear regression with the ridge regression predictor as regressor.