Advances in Business Statistics, Methods and Data Collection

28Sample Coordination Methods and Systems for Establishment Surveys

Alina Matei¹ and Paul A. Smith²

¹Institute of Statistics, University of Neuchâtel, Neuchâtel, Switzerland

²S3RI and Department of Social Statistics & Demography, University of Southampton, Southampton, United Kingdom

28.1 Introduction

Establishment statistics are important tools for economic management, and the need to obtain information as the basis for such statistics has led to legislation in many countries making response to establishment surveys compulsory. The converse of this requirement has been the management of survey burdens, a concept articulated by Bradburn (1978) so that the information requirement from government is limited. Burden management has manifested in two ways. First, through the reduction of numbers of surveys, sample sizes, and numbers of questions, all aiming to reduce the overall burden; and second, through procedures to spread the burden so that any individual establishment is selected in as few surveys as possible at a given moment. A major tool in this second approach has been the use of sample coordination approaches.

Sample coordination refers to methods used to optimize the overlap of two or more samples. The samples can be selected on different occasions in a repeated survey, or over different surveys. If a survey's objective is to estimate change across time, or to reduce the costs associated with recruiting a new sample unit, the overlap between samples should be maximized. ...