Nonlinear numerical models for the engineering design of mechanical connections in anisotropic materials require nonlinear material behavior of their components, which are essentially determined by material or structural testing. Herein, a multi-step approach for the parameterization of the nonlinear and anisotropic connection slip behavior is presented and applied to the ductile embedment behavior of steel dowels in wood. For this purpose, previously proposed regression functions for the slip behavior are reviewed, and further possible equations are discussed. Their suitability in the description of typical shapes of slip curves observed in connection testing is assessed before certain combinations are applied to an experimental dataset of embedment tests of steel dowels embedded in Laminated Veneer Lumber. The dependence of the regression parameters on the displacement range in the experimental dataset and the benefit of using parameters with a physical interpretation for being able to exploit connection test data reported in literature is highlighted.