A Practical Consideration on Proportionate Sample Design

Abstract

Sampling has been one of the powerful tools for researching any social phenomenon. Most of the sample surveys has multiple objectives and aims at providing separate estimates for different domain of the study. Accordingly, various domain of study is formed and attempt is made to represent them in the sample in different ways. Among them, proportionate design is the most preferred design. Equal probability selection method (epsem), representativeness and equal sampling rate are the basic features of the proportionate design. In general, a sampler has to deal with varying size of domain and clusters and choose appropriate method of sample selection in order to retain basic features of proportionate design. In this context, this article tries to demonstrate how appropriate choice of sampling methods can be made in relation to varying size of domains and clusters, and different types of sample design adopted.

The Economic Journal of Nepal, Vol. 35, No. 2, April-June 2012 (Issue N0. 138)