Abstract:
|
It is difficult to design a survey because prior information on response rates and the like is likely generated from a different random process than the target one governing the survey to be designed, and the impact on the properties of the estimator can be significant. Nowadays, computer-assisted data collection methods provide an instant variety of observations on the target random process governing the survey under consideration. These data and paradata enable the survey producer to make decisions regarding the need for methodology-process revision, which involves consideration of both a model that represents how the target information relates to the prior information and the design that describes how the observations are obtained. We think of the prior and target information as a random process that has a joint distribution with some probability function. Then at each phase of data collection, after receiving the information that the target random process has taken specific values, we update the joint probability distribution, to revise the design specification in the course of the data collection period. A coefficient of reliability is also discussed.
|