Reducing attrition in phone-based panel surveys: best practices and semi-automation for survey workflows

Abstract Panel surveys and phone-based data collection are essential for survey research and are often used together due to the practical advantages of conducting repeated interviews over the phone. These tools are particularly critical for research in dynamic or high-risk settings, as highlighted by researchers’ responses to the COVID-19 pandemic. However, preventing high attrition is a major challenge in panel surveys. Current solutions in political science focus on statistical fixes to address attrition ex-post but often overlook a preferable solution: minimizing attrition in the first place. Building on a review of political science panel studies and established best practices, we propose a framework to reduce attrition and introduce an online platform to facilitate the logistics of survey implementation. The web application semi-automates survey call scheduling and enumerator workflows, helping to reduce panel attrition, improve data quality, and minimize enumerator errors. Using this framework in a panel study of Syrian refugees in Lebanon, we maintained participant retention at 63 percent four and a half years after the baseline survey. We provide guidelines for researchers to report panel studies transparently and describe their designs in detail.