In this new era of "caller-id," voice mail screenings, declining landline penetration, ubiquitous cell phones, and constant telemarketing, the use of cold telephone calling based on conventional random-digit-dialing is becoming less effective as a technique for reaching random samples of a community's residents. It is important for us to revise and expand the sampling procedures for this new study.
The research team has reviewed the wide range of alternative survey methodologies being used in comparable survey projects across the nation. Also, we discussed the issue with several survey methodologists at the American Association for Public Opinion Research (AAPOR) Annual Conferece.
Given that a growing number of Americans are eliminating the traditional landlines in favor of cellphones, a trend led by the young and the poor, it is critical to secure interviews from a random-digit-dialing (RDD) sample of cellphone numbers. Broadly speaking, the sampling design for this study will include 70% of landline and 30% of cellular telephone numbers selected randomly from the pool of numbers expected to serve the Harris County area. We hope that the use of a dual-frame RDD sampling design may compensate for undercoverage bias.
Upon final approval of the questionnaires, in both English and Spanish, the data collection vendor will format and program the survey into the computer-assisted telephone interviewing system, also known as CATI, for use in interviewing.
- Last updated on August 1, 2012.