Skip to main content

Use of face coverings by the public during the COVID-19 pandemic: an observational study

Our take —

This study, a preprint and thus not yet available for peer review, presented direct observation of 3,271 patrons exiting grocery stores in 20 Wisconsin counties. It revealed fewer than half of shoppers donned face coverings in public with substantial variability in face covering use by county, gender expression, and age. Given that this study used a convenience sampling strategy, only included grocery stores that did not have face covering requirements, and may have incorrectly measured face covering use, these findings have limited generalizability to other public settings and in other US settings.

Study design


Study population and setting

Between May 16 and June 1, 2020, investigators used direct observations of individuals leaving 26 grocery stores (none of which required face coverings for entry) across 20 Wisconsin counties to determine the prevalence of face mask use in public settings. Observers documented shoppers’ visible age, gender expression, and the presence/type of face covering used. Based on field notes, the authors conducted a cross-sectional statistical analysis to determine if individual (i.e., age, gender expression), compositional (i.e., grocery store price index), and contextual (i.e., county population, COVID-19 case prevalence) factors were associated with use of face coverings.

Summary of Main Findings

Across 3,271 observations, 41.2% donned face coverings in the observation period, with substantial heterogeneity across counties (range: 6.8% to 69.0%). In regression analysis, face mask use was significantly higher among adults (CI: 1.06 – 2.07) and older adults (CI: 2.86 – 5.96), compared to minors. Women were more likely to have face coverings than men (CI: 1.36 – 1.86). Individuals had higher odds of observed face coverings in stores with higher price indices, as determined through the price of 12 staple foods (CI: 1.34 – 1.86). In four of the five most populous counties, higher COVID-19 case prevalence estimates correlated with lower proportions of individuals observed with face coverings.

Study Strengths

This study had a moderately large sample size with significant variability in face coverage usage in the US population.


Deriving face covering behavior from a convenience sample of grocery store patrons has limited generalizability, as shoppers’ use of face coverings in a grocery store may not align with their use of face coverings in other public settings. Furthermore, the convenience sample may not represent all grocery store patrons in Wisconsin. The validity of the observations on face covering usage is also questionable, given that observed shoppers might have removed face coverings prior to exiting grocery stores.

Value added

The study offers new evidence of use of face coverings in public settings in the United States through direct observation.

This review was posted on: 10 July 2020