Abstract

The Annual Survey of Establishments (ASE) is a national survey of the National Statistics Office that provides detailed information on employment, salaries and wages, capital expenditures, outputs and sales, inventories and other characteristics of establishments. These data assist in assessing the structures, level and growth potentials of establishments by economic sector, size, and region on a time series. Inspite of its regularity as a data base, ASE apparently suffers from non-sampling errors (NSE). Recommendation of the study dealt on NSE on: (1) frame/coverage concepts 9defining role additional samples, developing system for detecting shifts of establishments, investigating misclassification and choice of stratification for shifts, (2) non-response (defining unit and item nonresponse, systematic recording of closed and cannot be located establishments, studying effects of notices to responses); and (3) processing (evaluating quality of processing, reflecting updates such as imputations and verifications in the questionnaire).