On testing exponentiality under Type-I censoring

Abstract

Two new goodness-of-fit testing procedures are introduced to test exponentiality when data are subject to Type-I censoring. We proposed four test statistics for this purpose. Under extensive Monte Carlo simulations, we showed that the proposed tests maintain the nominal significance level and show good power for both monotonic and non-monotonic hazard function alternatives even for small samples as n = 10. A real dataset is studied for illustrative purposes.