作者: JOHN J. GART , JAMES R. ZWEIFEL
DOI: 10.1093/BIOMET/54.1-2.181
关键词:
摘要: SUMMARY The bias of several logit estimators and their corresponding variance is investigated in small samples. Their use quantal bioassay similarly explored. transformation has been suggested the analysis higher dimensional contingency tables, by Woolf (1954) many others more recently, also estimating parameters logistic function problem (Berkson, 1944, 1953). Various modifications have Berkson (1953), Haldane (1955), Anscombe (1956), Tukey, mentioned Hitchcock (1962). Modifications its usual estimator proposed Haldane, Goodman (1964) Gart (1966). More recently Goodman, an unpublished paper, derived further his estimator. In this paper we present a numerical comparison these give conditions under which one or other may be preferred. logits briefly explored with particular reference to asymptotic results regarding estimators.