Changing Modes on the Fly: Transitioning a Complex Longitudinal Survey from In-Person to Phone Due to COVID-19
Becky Reimer
NORC at the University of Chicago
Kylie Carpenter
NORC at the University of Chicago
Ann Bisognano
NORC at the University of Chicago
Liz Kantor
NORC at the University of Chicago
Due to COVID-19, many longitudinal survey projects were forced to change course without substantial planning or field testing. Among them was the Medicare Current Beneficiary Survey (MCBS), an in-person survey of a nationally representative sample of the Medicare population, conducted by the Centers for Medicare & Medicaid Services (CMS) through a contract with NORC at the University of Chicago. The MCBS transitioned to phone administration in March 2020 after a brief pilot test phase in production. The present study investigates the data quality impact of the transition, focusing on reporting of health care events and associated costs, which became more burdensome via phone because interviewers were unable to manually review documents (e.g., statements) as they did prior to the transition. Many techniques included in the mode analysis literature do not apply to this work due to the lack of experimental design and synchronous data collection across modes. Instead, we assess the degree to which the data collected via phone in 2020 continued or broke existing trends. We discuss difficulties separating the impact of the mode change from the impact of the pandemic itself.