This chapter treats the basic methods for connecting the model to empirical data: estimating and testing parameters. These are based on computer simulations of the model with provisional hypothetical values of the model’s parameters. To give some intuitive understanding, the chapter starts with a toy model. It goes on to describe how parameters are estimated in the RSiena package by the method of moments and the Robbins–Monro algorithm. Subsequently, methods are given for testing the parameters. Special attention is given to the score-type test, which is a method for testing parameters without estimating them. This can be useful, for example, for cases of complete or quasi-complete separation, where estimated parameter values would be infinitely large or nearly so. Conceptual questions about whether it is meaningful at all to use statistical models for data are discussed in a section about model-based inference. The chapter ends with a brief history of this methodology for analyzing network dynamics.