New models have been developed that accurately describe the response surfaces of capacity factors that are a function of changes in the pH and the fraction of organic modifier in reversed-phase high-performance liquid chromatography (RP-HPLC). The purpose of this article is to illustrate one of the problems of nonlinear modeling of capacity factors; it is an extension of the articles on this nonlinear modeling in RP-HPLC that have appeared during the last three years. The main subject of this paper is a description of the methods that can be used in building correct models for the capacity factor as a function of both pH and modifier content of the mobile phase. The heterogeneity of the variance across all capacity factors causes the ordinary nonlinear least squares method to be less appropriate for regression. Two methods are described that take account of the effect of the heterogenic error structure of capacity factors in the modeling procedure. These two methods are then compared with the standard nonlinear modeling procedure.
