Estimating the difference between two means with missing data in sample surveys

Article type: Research Article

Authors: Rueda, M.^{a; *} | González, S.^b | Arcos, A.^a

Affiliations: [a] Department of Statistics and Operational Research, University of Granada, Spain | [b] Department of Statistics and Operational Research, University of Jaén, Spain

Correspondence: [*] Corresponding author: Department de Estadística e I.O., Facultad de Ciencias, Avda. Fuentenueva, Universidad de Granada, 18071, Granada, España. Tel.: +34 95824 0494; Fax: +34 95824 3267; E-mail: [email protected]

Abstract: The objective of the present study is to estimate the difference between two population means when some observations on the first and second characteristics are missing form the sample. The proposed estimator defined for general sampling design is computationally simple and an easy expression for its variance is obtained. The proposed estimator is applied to two years data concerning the commercial operations in Andalucia.

Keywords: Auxiliary information, missing data, Horvitz-Thompson estimator, simple random sampling

DOI: 10.3233/MAS-2006-1109

Journal: Model Assisted Statistics and Applications, vol. 1, no. 1, pp. 51-56, 2006