Abstract
Missing data due to nonresponse, though undesirable, is a reality of any survey. In this paper we consider a situation in which, at a given time, observations are missing for one of the several auxiliary characteristics; thus the ‘missing’ phenomenon occurs for the characteristics separately but not simultaneously. A new method, making use of all the available observations, is proposed. A simulation study based on three real populations was performed to test the proposed technique.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Little RJA, Rubin DB (1987) Statistical analysis with missing data. Wiley, New York
Rao JNK (1994) Estimating totals and distribution functions using auxiliary information at the estimation stage. J Official Stat 10(2):153–165
Rueda M, González S (2004) Missing data and auxiliary information in surveys. Comput Stat 19:551–567
Singh S, Joarder AH (1998) Estimation of finite population variance using nonresponse in survey sampling. Metrika 47:241–249
Toutenburg H, Srivastava VK (1998) Estimation of ratio of population means in survey sampling when some observations are missing. Metrika 48:177–187
Tracy DS, Osahan SS (1994) Random nonresponse on study variable versus on study as well as auxiliary variables. Statistica 54:163–168
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
González, S., Rueda, M. & Arcos, A. An improved estimator to analyse missing data. Stat Papers 49, 791–796 (2008). https://doi.org/10.1007/s00362-007-0045-8
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00362-007-0045-8