ABSTRACT
The conduct of large scale surveys in national statistical offices follows a wellestablished process that involves a large number of steps. These steps, which are described in Statistics Canada’s Quality Guidelines (Statistics Canada, 2008), begin with the setting of objectives for the survey and end with dissemination and documentation. In this chapter, we describe some of the important developments that have taken place at Statistics Canada to improve the intermediate steps. Some of the steps that we will describe include record linkage, sample design, editing, imputation, weighting, estimation, variance estimation, data analysis, benchmarking, and the use of time series. Many of these developments were necessary to improve the quality of the resulting published estimates, to reduce costs, and to aid in the building of either specific or generalized software to automate the methodology.
