Estimating The Number And Location Of Knots In Spline Regressions

Abstract

The purpose of this article is to demonstrate a simple method of estimating the number and location of knots (join points) in spline regressions.

 Spline regression models offer a convenience alternative to dummy (binary) variable models.  Using a dummy variable to alter the intercept or slope of a model generally results in a break in the regression line.  For example, a substantial drop in a firm’s production may occur if it suddenly closes a major plant.