Focused model selection for social networks

•Proposing a focused parameter selection method for social network models.•Focused selection is performed for various classes of network models.•The procedure is driven by a main parameter of interest called the focus.•The mean squared error value for the estimator of the focus is estimated.•Several relevant focuses are used throughout to exemplify the method. We present a focused selection method for social networks. The procedure is driven by a focus, the main quantity we want to estimate well. It represents the statistical translation of a research hypothesis into parameters of interest. Given a collection of models, the procedure estimates for each model the mean squared error of the estimator of the focus. The model with the smallest such value is selected. We present focused model selection for (i) exponential random graph models, (ii) network autocorrelation models and (iii) network regression models to investigate existing relations in social networks. Worked-out examples illustrate the methodology.