Purpose:To present the computational tools called indicators for the selection of auxiliary variables based on their effectiveness.Summary:To control nonresponse bias in the estimators the most suitable approach is to use auxiliary variable information. In many Scandinavian countries such information is available in administrative registers. Using such information in surveys effectively is called design for estimation and methods for this are developed and used in Sweden in many surveys. The process of selection of the auxiliary variables is discussed in detail in many recent studies. Like selection of auxiliary variables, their weighing is also equally important. There are two approaches considered for weighing in the studies namely providing a general setting and quantifying the search with indicators. The article extends the second approach in the selection of auxiliary variables. (10 refs.)
展开▼
