|
For business surveys at Statistics Canada, there are significant costs to the follow-up operation in data collection. In order to improve cost efficiency under the constraints of a limited budget, different methods have been employed in data collection for many business surveys at Statistics Canada.
In this paper, a simulation-based approach is developed for comparing different methods and determining optimal ones for the follow-up operation. The follow-up operation environment is simulated using different scenarios based on the Reference Year 2006 data obtained from the Unified Enterprise Survey at Statistics Canada. The different collection methods are applied to the data at certain levels of domain of interest and are then compared by measuring their impacts on the final estimates of economic variables as well as on the cost associated with each collection method.
|
