This administrative supplement had continued a high-frequency longitudinal survey of Americans’ experiences and behavior during the COVID-19 pandemic. The survey had been conducted with respondents to the Understanding America Study (UAS), a probability-based panel of 8,500 adults representing the entire United States.

Housed in the Center for Economic and Social Research at the University of Southern California, the UAS had employed an ‘Internet Panel,’ in which respondents answered surveys on a computer, tablet, or smartphone, wherever they were and whenever they wished to participate. Respondents had been recruited through Address-Based Sampling and had received a tablet and broadband Internet subscription if needed, thus facilitating coverage of the entire adult population 18 and over in the U.S. The set-up had allowed for an immediate and efficient transmission of data, which had been quickly made publicly available through its online platform.

The team had invited 7,000 UAS respondents to answer bi-weekly surveys (500 every day) through the rest of the year. They had reported moving weekly averages that were updated every night by incorporating the newest batch of responses. Thus, results had been based on a rotating sample of responses. Importantly, since the same respondents had been answering every other week, the team had been able to track changes with much more accuracy than when new samples were drawn every week. Updated results had been posted on the UAS website every night.