With cell-phone sampling increasingly used to supplement traditional landline telephone surveys as a means of improving sample coverage, debate continues regarding the best approach to including the cell-phone frame. One school of thought suggests that households should be screened for landline usage. The screening approach may seek to identify households that are cell-phone-only (CPO), “cell mostly” (Blumberg et al.), “cell mainly” (Boyle et al.), or some combination of categories. The goal in screening is to spend cell phone interviewing effort on only those households that are not included in the landline frame. But there are imperfections inherent in the screening approach and further research by Boyle et al. (2011) suggests that a screening approach may yield a demographic distribution of the sample that is further from benchmark distributions than a take-all approach, where all eligible cell telephone households are interviewed regardless of telephone usage. Further, the screening approach can increase costs, although this may be offset by the reduction in effective sample size that the design effect of an unscreened sample may incur.

The purpose of this work is to examine the demographic distributions obtained in dual-frame telephone surveys in which the cell-phone interviews are conducted under a take-all approach versus a screening approach. Using data from the Associated Press-NORC Center for Public Affairs Research – Reflections on 9.11 study – we seek to determine which of the two approaches yields distributions of demographics and phone status that are closer to benchmark distributions. We construct the distributions selectively using dual-users and CPO cases from the cell sample and all landline cases from the landline sample. Distributions based on the take-all approach include all dual users from the cell sample and the landline sample, while distributions based on the screening approach are simulated by removing respondents who are not CPO from the cell sample. We compare these results to external benchmarks on key demographic measures, and also consider the cost/design effect tradeoff . We also consider using cell-mostly and cell-mainly cases as part of a screening design and compare our results to those found in prior studies.