An introduction to nonparametric regression in demographic research

Nonparametric (NP) regression methods are increasingly important in statistics, economics, and other disciplines. To date, however, very little demographic research has made use of NP methods. These new methods utilize increased computing power to limit the need for time-saving (but possibly incorrect) mathematical assumptions about functional forms in regression analysis. This expository, non-technical paper discusses the rationale for NP regression, illustrates several of the basic NP regression methods with demographic examples, describes extensions of basic methods to multivariate regressions, and discusses the relevance of NP regression for demographic research.