Using meta-analysis for data enrichment-optimal families of estimation strategies. | [Journal of Applied Sciences • 2005]

Author(s):

1. Ashok Sahai: Department of Statistics and Demography, University of Swaziland, Kwaluseni, P. Bag 4, Swaziland, Africa

2. M. Mahibbur Rahman: Department of Statistics and Demography, University of Swaziland, Kwaluseni, P. Bag 4, Swaziland, Africa

3. Rameshwar P. Jaju: Department of Computer Science, University of Swaziland, Kwaluseni, P. Bag 4, Swaziland, Africa

Abstract:

Long-term data-sets of good quality are invaluable in certain research investigations. In some situation, they are so expensive of time and money, that they are seldom available. It has been noted that the problems of varying data quality, of missing information and that of diversity of data might possibly be dealt with by using meta-data. The key point is to treat relevant good quality meta-data as auxiliary/ancillary information, the use of which is gained through the proposed optimal simple estimation strategies. These strategies are motivated by mixing-type estimators, the desired optimality of these estimators is achieved through optimal manipulation of two design-parameters therein set to control first and second order of large sample approximations to their standard errors. A sensitivity analysis is then used to discover, empirically, the robust estimation strategy from amongst the proposed alternatives.

Page(s): 1575-1581

DOI: DOI not available

Published: Journal: Journal of Applied Sciences, Volume: 5, Issue: 9, Year: 2005

Keywords:

Keywords are not available for this article.

References:

References are not available for this document.

Citations

Citations are not available for this document.

Citations

Downloads

Views