ABSTRACT
We have been working with linear (in the parameters) regression equations in which the regressors involved only fi rst-degree terms, i.e., for either the univariate (Y = β0 + β1X + ε) or multivariate (Y = β0 + β1X1 + … + βmXm + ε) cases, each regressor appears only to the fi rst power. In general, as we shall soon see, the degree of a polynomial model is the maximum power of any term in the model, whereas the order of the model is the maximum sum of powers for any individual term. The name given to a polynomial model refl ects the highest degree (or largest power) of any single regressor. Thus, for univariate polynomial models, their degree always equals their order.
