Can a Statistician Deliver?

Abstract

The theme in this article is quality in the statistics produced by National Statistical Agencies. Good survey methodology is important for realizing quality. Survey methodology has been described as a collection of practices backed by some theory. It is perhaps too optimistic to expect a complete and coherent theory for as vast a field of activities as national statistics production. Yet the desire for a more satisfactory theoretical basis is voiced by many. We discuss progress and failures in survey theory advancement. We note that the different types of statisticians participating in the production process may view quality in possibly different ways; their views, on the other hand, may not wholly agree with those held by the users of statistics.