AAPOR2013 - Address-based Sampling (ABS) as an “Alternative” to RDD: A Test in California - Procedural Results

Sep 13, 2026

Address-based sampling (ABS) using the USPS Delivery Sequence File (DSF) is evaluated as an alternative to landline random-digit dial (RDD) sampling to address declining coverage in landline frames while reducing costs versus dual-frame cell/landline RDD. DSF coverage is stronger in urban areas, and ABS enables contact and response options suited to hard-to-survey groups, including ethnic and linguistic minorities and households with only a cell phone. A California Health Interview Survey pilot applied mailing-based screening and telephone interviewing, reporting procedural findings from sampled communities.

LanguageEnglish

CategoryPresentations

FormatPDF

Pages10

Reading Time4 min

AccessFree to download

Abstract

1. Introduction

Motivation for ABS

Pilot research question

2. Methods

2.1 Sample Design

AAPOR2013 Address-based Sampling (ABS) as an “Alternative” to RDD: A Test in California Matt Jans 1 , David Grant 1 , Annie Lee 1 , Royce Park 1 , Sherman Edwards2 , John Rauch2 , Ismael Flores-Cervantes2 1UCLA Center for Health Policy Research, California Health Interview Survey, 10960 Wilshire Blvd., Ste. 1550, Los Angeles, CA 90034 2Westat, 1600 Research Boulevard, Rockville, MD 20850-3129 Abstract Address-based sampling (ABS) from the USPS Delivery Sequence File (DSF) presents a potential method to overcome historical coverage decreases in landline random digit dial (RDD) frames, and reduce costs relative to dual-frame cell/landline RDD samples. DSF coverage tends to be better in urban areas than rural areas, and the contact and response options available with ABS make it an interesting method for reaching otherwise “hardto-survey” populations, such as ethnic and linguistic minorities, and households with only a cell phone. ABS was pilot tested as a special project of the California Health Interview Survey (CHIS), and the procedural results are presented here. Two California communities comprised the target population (with a sample of n=8,277 addresses) . Communities where chosen based on population size/density, ethnicity, and percentage speaking Spanish. The mailing protocol included three "full-packet" mailings of a onepage screener form designed to replace screening traditionally done via phone. A reminder postcard was mailed between the first and second mailings. Households providing a phone number were called to complete the standard CHIS telephone interview. Households not providing a phone number were called if one was matched to their address through a record match. Initial results suggest that ABS may work as a replacement to RDD screening but with several caveats. Key Words: address-based sampling, ABS, hard-to-reach populations, RDD 1. Introduction Landline random-digit dial (LL-RDD) surveys have decreased precipitously in coverage of the general household population over the past decade as people and households rely more on cell phones (Blumberg & Luke, 2007, 2010; Blumberg et al., 2011 , 2012) . The California Health Interview Survey (CHIS) has moved to a dual-frame LL/cell-RDD sample design to reduce coverage error related to this trend. Yet, adding a second sampling frame increases the complexity of sampling, data collection, and weighting. Further, precise geographic targeting is difficult with cell-RDD samples because phone numbers and the phones themselves are not tied to a specific geographic location as reliably as landline numbers are. Thus, sampling households by sampling addresses has benefits over RDD sampling for surveys of small geographic areas. Further, addressbased sampling (ABS) may lead to more efficient data collection if households can be contacted and screened by mail instead of telephone because no interviewers are involved AAPOR2013 and geographic screening will be more efficient (i.e., fewer sample units that are not in the target area) . When considering an ABS design, phone surveys like CHIS need to decide whether to use an ABS approach for contact and screening only or for data collection as well. Several factors make transitioning large phone surveys like CHIS to a mail mode. CHISis conducted in five languages, has complex skip patterns, and interviews multiple randomly-selected people in each household, each of which makes computerized interviewer-administered options a better choice. The research question for this pilot test was “Can we (and how do we) use a mail mode to obtain phone numbers at which to call sampled addresses?” The ABS pilot test reported here has the following goals: 1) Test the logistics and feasibility of using an ABS sample to screen households and obtain phone numbers by mail and computer match, but conduct interviews by phone. 2) Evaluate how well the cell phone only (CPO) population can be reached by this sequential multi-mod

FAQs

Why is ABS considered an alternative to landline RDD surveys?

ABS uses USPS DSF to counter declining landline coverage, potentially improve reach in targeted areas, and reduce costs compared with dual-frame cell/landline RDD designs.

How was the ABS pilot implemented in the California Health Interview Survey?

The pilot sampled 8,277 addresses in two communities, used full-packet mailings with a one-page screener form, sent a reminder postcard, and then conducted telephone interviews using matched phone numbers or phone numbers provided by households.

What were the main goals and comparisons of the pilot test?

The pilot tested logistics and feasibility of mail-based screening and phone number acquisition, assessed reachability of the cell-phone-only population, and compared results across communities with different Spanish-speaking population proportions and other demographic factors.

Sep 13, 2026

Uploaded by Olivia Brown