This study aims to obtain tables of critical values of the modified Kolmogorov-Smirnov, Anderson-Darling and Cramér-von Mises goodness of fit tests for the flexible Weibull distribution with unknown two parameters in case of complete and type-II right censored samples. The power study of these tests will conduct for several alternative distributions. Meanwhile, the power comparisons among these modified tests, a goodness of fit test based on data transformations which have been introduced by Goldmann et al. [8] and the exact goodness of fit test with type-II right censored samples which has been suggested by Grané [9] are investigated for a number of alternative distributions. Moreover, we use real data set as an illustration example for goodness of fit tests for the flexible Weibull distribution.

goodness of fit, censored samples, flexible Weibull, Monte Carlo simulation.