Tests for the Exponential, Weibull and Gumbel Distributions Based on the Stabilized Probability Plot
研究了用于指数、威布尔和冈贝尔分布复合检验的拟合优度方法，这些方法可与概率图结合使用，且比科尔莫戈罗夫-斯米尔诺夫检验更有效。
Goodness-of-fit procedures for composite testing of the exponential, Weibull and Gumbel distributions are investigated. These procedures are such that they can be used graphically with, and to aid interpretation of, probability plots. In the situations of interest here, the new procedure is generally more powerful than the Kolmogorov- Smirnov test, which appears to be the only other readily available goodness-of-fit procedure that can be used graphically with probability plots.