The jackknife resampling method has been successfully applied to many complex inference problems. At its most basic, this method involves systematically removing one observation from the data and estimating parameter(s) in the resulting sample, and then recombining these estimates. In this study, we consider a “jackknife” type procedure to network modelling by repeatedly obtaining parameter estimates using logistic regression while randomly deleting one edge from the graph network. Our goal is to determine how well the predicted networks from the model fit the observed network through goodness of fit measures. This method allows us to also comment on the robustness of this method.