Logo Logo
Hilfe
Hilfe
Switch Language to English

Schwaiger, Manfred ORCID logoORCID: https://orcid.org/0000-0003-0132-4560 (1995): Improving Sample Results by Object-Weighting. International Conference on Operations Research, Berlin, August 30 – September 2, 1994, Berlin, August 30 – September 2, 1994. Derigs, Ulrich; Bachem, Achim und Drexl, Andreas (Hrsg.): In: Operations Research Proceedings 1994, Berlin, Heidelberg: Springer Berlin Heidelberg. S. 541-546

Volltext auf 'Open Access LMU' nicht verfügbar.

Abstract

This article brings a parameter estimation procedure into focus that takes bias into account by calculating specific weights for each sample object. A sample consisting of survey-variables, whose expected values in the population are to be estimated, and auxiliary variables, whose mean values are known in the population, serves as a database. The sample is biased, if the mean values of the auxiliary variables differ from the respective mean values in the population. In this case classical estimation provides biased point estimators. Empirical results show that weighted estimation offers more precise results than classical estimation does.

Dokument bearbeiten Dokument bearbeiten