A Reassessment of List-Assisted Rdd Methodology

Abstract

Random digit dial (RDD) sampling methodology was devel- oped over two decades ago when local telephone exchanges used 100- series telephone banks as physical building blocks for telephone number assignment. During the intervening decades the telecommunication in- dustry has undergone a number of fundamental changes that have had drastic effects on the efficiency and coverage of RDD samples. Based on a new study of landline telephone numbers, which was conducted during the second quarter of 2008, this paper reexamines the assumptions that are relied upon for construction of list-assisted RDD samples; quantifies the extent of undercoverage in the corresponding sampling frames; in- vestigates alternative methods of frame construction for RDD samples; and evaluates other sample design options that can offer greater coverage with varying degrees of efficiency.