The log-linear model with a generalized gamma distribution for the error: A Bayesian approach (Q1821445)

Considering a log-linear model with one covariate and a generalized gamma distribution for the error, we find the posterior densities for the parameters of interest. Since many standard survival distributions are particular cases of the generalized gamma model, the proposed Bayesian method is very useful to discriminate between possible models to be used in the data analysis. The Laplace approximation for integrals [see \textit{L. Tierney} and \textit{J. B. Kadane}, J. Am. Stat. Assoc. 81, 82-86 (1986; Zbl 0587.62067)] is used to find the posterior distributions of the parameters involved when they cannot be obtained explicitly.