Composite estimation is a statistical estimation procedure that combines data from several sources, for example, from different surveys or databases or from different periods of time in the same longitudinal survey. It is difficult to describe the method in general, as there is no limit to the ways one might combine data when various useful sources are available. Composite estimation can be used when a survey is conducted using a rotating panel design with the goal of producing population estimates for each point or many points in time. If the design incorporates rotating groups, composite estimation can often reduce the variance estimates of level variables (e.g. totals, means, proportions). In addition, composite estimation can reduce the variance estimates of variables dealing with changes over time, ...
Looks like you do not have access to this content.