Book Description

Extract: Multiple-frame surveys are susceptible to errors that stem from associating the over-lapping portions of the frames. Decreasing sampling error may improve precision, but it offers greater complications; nonsampling error would increase, leading to decreased accuracy and, perhaps, greater total error. This manuscript offers several procedures to improve consistency and accuracy of multiple-frame estimating by reducing nonsampling errors. Such procedures, when added to an optimum mix of sampling frames, would provide survey indications of greater reliability without increasing costs appreciably.