Dear all, I am trying to get a more accurate estimate of the dispersion parameter of the inverse.gaussian family in a glm model. The one provided by the summary.glm looks like only a rough estimate, when you calculate the individual likelihoods and sum them using the dispersion reported by summary.glm, it can get quite different from the reported log-likelihood value. As discussed in a previous thread, for the Gamma family, there is gamma.shape function that does this more accurately. Is there a counterpart function that calculates ML estimator of the dispersion for the inverse.gaussian family? Thanks, Alex Pegucci