THE EFFECT OF BIAS, VARIANCE-ESTIMATION, SKEWNESS AND KURTOSIS OF THE EMPIRICAL LOGIT ON WEIGHTED LEAST-SQUARES ANALYSES
成果类型：
Article
署名作者：
GART, JJ; PETTIGREW, HM; THOMAS, DG
署名单位：
National Institutes of Health (NIH) - USA; NIH National Cancer Institute (NCI)
刊物名称：
BIOMETRIKA
ISSN/ISSBN：
0006-3444
DOI：
10.2307/2336348
发表日期：
1985
页码：
179190
关键词：
摘要：
The bias and first 4 cumulants of the empirical logit tranformation of a binomial variate are studied by means of asymptotic expansions and exact computation. The covariance of the empirical logit and its estimated variance is derived. Neither the +1/2 correction of Haldane (1955) and Anscombe (1956) nor the -1/2 suggested by Cox (1970) for weighted logit regression is universally effective in reducing bias. Other corrections are considered. The distribution of the empirical logit is considerably more skewed and has a much larger kurtosis than the comparable binomial distribution. Methods based directly on sufficient statistics are preferred to those based on the empirical logit.