We propose in this paper the use of zero-inflated Geometric model, which is a generalization of the Geometric model, in the analysis of reliability and survival data. The use of this model is necessary especially when the data presents an excessive number of zeros. Maximum likelihood estimates of parameters were obtained, even as their confidence intervals based on asymptotic theory. In addition, we use the bootstrap resampling technique as an appropriate alternative procedure to construct confidence intervals of parameters of zero-inflated Geometric model.
Intervalar estimation; coverage probability; bootstrapping technique
