Journal of Official Statistics, Vol.18, No.2, 2002. pp. 129–154
Models in the Practice of Survey Sampling (Revisited)
Abstract:Although design-based inference is the standard form of inference with large-scale sample surveys, in practice some reliance on model-dependent inference is necessary. This article considers issues of model-assisted inference, the population of inference, conditional inference, the effect of measurement errors, and analytic uses of survey data, under the design-based mode of inference. It then discusses the need for model-dependent inference for small area estimation and for handling missing data (unit nonresponse, item nonresponse, and noncoverage). The article also discusses the use of models in variance estimation.
Keywords:Design-based inference; model-dependent inference; missing data; small area estimation; generalized variance functions.
Copyright © Statistics Sweden 1996-2018. Open AccessISSN 0282-423XCreated and Maintained by OKS Group