A Method of Increasing Power of a Test for the Negative Binomial and Neyman Type A Distributions
修改广义最小卡方过程中的二次型，得到检验统计量在非零分布下的新形式，用于检验负二项分布和内曼A型分布的拟合优度，并给出渐近功效表显示功效提升。
Abstract A modification of the quadratic form involved in a generalized minimum chi-squared procedure yields the same asymptotic null distribution of the test statistic for testing fit by an assumed model, but a different form of nonnull distribution emerges. This result is applied to testing fit of the negative binomial and Neyman Type A distributions, and tables of asymptotic power are presented that show the increase of power obtained. A numerical example is presented to illustrate the technique. Key Words: Goodness of fitModified testTest statisticNonnull distributionChi-squared distributionAsymptotic power