Description
A response burden arises from the need for statistical information about finite populations. In business surveys the response burden is an important aspect since one business may fall in surveys many times in a given time interval. This raises a question. How to even out this burden as fairly as possible? Poisson sampling, with Bernoulli sampling and Bernoullis sampling and strict Poisson sampling as special cases, has been found in the early seventies to have good properties as regards sample co-ordination. A new approach for sample co-ordination, the Poisson Mixture (PoMix) sampling is introduced in this study. This is a sampling scheme which partly uses Bernoulli sampling scheme and partly Strict Poisson sampling scheme. This study also proves this sampling scheme to be more efficient than the traditional Poisson Sampling. Ismo Teikari is Senior Researcher in Statistics Finland. The resipient of DSc.(Econ) from the Helsinki School of Economics and Business Administration (2002). Expertise in business surveys, sampling co-ordination and empirical micro econometrics for business data.
