Smoothing Splines Application in Comparison Process

The theory of smoothing splines in the context of semi-parametric generalized linear models is well developed (Green and Silverman 2008), but applications tend to raised computational and pragmatic problems. In this paper an example is considered and we compare the use of smoothing splines in logistic regression model, having discretized event history sequences both with and without random effect specification to control for omitted variables. The main emphasis is upon using cubic spline functions for a number of temporal variables.