Evaluating Substitution as a Strategy for Handling Drop Points in Self-Administered Address-Based Sampling Frame Surveys

• Activity Number: 526 - Trends in Sample Design
• Type: Contributed
• Date/Time: Thursday, August 11, 2022 : 8:30 AM to 10:20 AM
• Sponsor: Survey Research Methods Section
• Author(s): Taylor H Lewis* and Joseph McMichael and Charlotte Looby
• Companies: RTI International and RTI International and RTI International
• Keywords: Address-Based Sample; ABS; health survey; substitution

Abstract

Most addresses on modern address-based sampling frames derived from the United States Postal Service’s Computerized Delivery Sequence (CDS) file have a one-to-one relationship with a housing unit. Some addresses, however, are associated with multiple housing units. These addresses are referred to as drop points, and the housing units therein are referred to as drop-point units (DPUs). DPUs pose a challenge for self-administered surveys because no apartment number or unit designation is available, making it impossible to send targeted correspondence. Nationwide, approximately 1.5% of addresses are DPUs, but they are concentrated most heavily in New York, New Jersey, Massachusetts, and Illinois. For example, upwards of 14% of the addresses comprising the Healthy Chicago Survey’s target population are DPUs. This paper evaluates a method for substituting sampled DPUs with similar non-DPUs, which was implemented in the 2021 Healthy Chicago Survey alongside a concurrent survey of the originally sampled DPUs. We do so by comparing the DPUs and non-DPU substitutes on metrics such as response rates and aggregate distributions of demographics and key health outcomes.