Web Survey Bibliography
Title Evaluation of Differential Sampling in the Knowledge Networks Panel: What is the Effect on Coverage, Variance and Bias for Key Statistics?
Author Huggins, V. J., Tang, X.
Year 2002
Access date 14.06.2004
Abstract Sample selection for the Knowledge Networks Panel is accomplished using standard RDD methods, which, in principle, generates equal probability sample designs. To improve the efficiency of some estimates and to reduce cost, oversampling of certain groups is applied. The following groups are differentially sampled in the Knowledge Networks Panel: Households with black and Hispanic members, households for which addresses can be obtained for the randomly generated telephone numbers, households in central region states, and households in areas of the U.S. not serviced by the MSN(r)TV as an Internet Service Provider. Results of an evaluation of the impact of the differential sampling on cost, coverage, sampling variance, and bias for key statistics will be discussed.
Access/Direct link Homepage - conference (abstract)
Year of publication2002
Bibliographic typeConference proceedings
