摘要
In many settings,multiple data collections and analyses on the same topic are summarised separately through statistical estimators of parameters and variances,and yet there are scientificreasons for sharing some statistical parameters across these different studies.This paper summarises what is known from large-sample theory about when estimators of a common structuralparameter from several independent samples can be combined functionally,or more specificallylinearly,to obtain an asymptotically efficient estimator from the combined sample.The main ideais that such combination can be done when the separate-sample nuisance parameters,if anyexist,vary freely and independently of one another.The issues are illustrated using data from amulti-centre lung cancer clinical trial.Examples are presented to show that separate estimatorscannot always be combined in this way,and that the functionally combined separate estimators may have low or 0 efficiency compared to the unified analysis that could be performed bypooling the datasets.
基金
The authors gratefully acknowledge the Eastern Cooperative Oncology Group as the source for the ECOG EST 1582dataset,and the suggestion of a referee to expand our treatment of(V)to estimating equations.
