AN EVALUATION OF PRIMARY DATA-COLLECTION MODES IN AN ADDRESS-BASED SAMPLING DESIGN

Area of Science:

• Survey Methodology
• Public Health Research
• Data Collection Optimization

Background:

• Address-based sampling is increasingly used in multimode surveys.
• Research is needed to optimize multimode designs across different data collection methods.
• Previous studies have not fully evaluated mode sequencing effects on mail and phone surveys.

Purpose of the Study:

• To evaluate two multimode case-flow designs: phone-first and mail-first.
• To compare response rates, cost, timeliness, and data quality between designs.
• To examine how eligibility rates influence the optimal case-flow design.

Main Methods:

• An experiment was conducted within the Racial and Ethnic Approaches to Community Health (REACH U.S.) survey.
• Two case-flow designs were tested: mail followed by phone (mail-first) and phone followed by mail (phone-first).
• Key metrics analyzed included response rates, cost-efficiency, data collection timeliness, and data quality.

Main Results:

• The mail-first design generally outperformed the phone-first design across most metrics.
• Mail-first achieved higher yield rates and lower costs with comparable data quality.
• Although phone-first initially yielded more interviews, mail-first ultimately secured a greater number of interviews over time.

Conclusions:

• The mail-first case-flow design is often superior for optimizing multimode surveys.
• This design offers a more cost-effective and efficient approach to maximizing survey participation.
• The choice of optimal case flow may be influenced by the target population's eligibility rate.