The Fisher Information Matrix on a Type II Doubly Censored Sample from a Dagum Distribution

We focus on the calculation of the Fisher information contained in a random sample with type II censored data from a Dagum distribution. Unlike the decomposition methods already known in the literature, we provide a direct calculation of the information matrix using one reparametrization of the density of the Dagum variable. The inverse of this matrix can be used to determine analytically the asymptotic variance-covariance matrix of the maximum likelihood estimates of the unknown parameters.