作者: James M. Lepkowski , Robert M. Groves
DOI: 10.1080/01621459.1986.10478352
关键词:
摘要: Abstract An error model for dual frame survey designs is developed. It includes components of sampling variance, interviewer and bias in each frame. A cost that attempts to capture the complexity a full scale presented. The models are applied large national survey, National Crime Survey, effect alternative levels both frames have on optimal allocation sample two examined types crime.