Abstract
Given a doubly censored sample from the generalized Rayleigh model we derive the likelihood function and the predictive densities for future responses. We consider several values for the parameters and provide different scenarios for the generalized Rayleigh model. The predictive densities are displayed with some assumed values of the parameters. The raw moments, corrected moments, skewness, and kurtosis for a future response are determined with some neighborhood values of the maximum likelihood estimates (MLEs) of the parameters. Two real data sets are used to obtain the predictive inference.