Study population and setting
Investigators implemented an internet survey with 503 adults in the U.S. between April 7 and 9, 2020—shortly following the CDC and U.S. White House COVID-19 Task Force’s joint recommendation of donning cloth face coverings in public (April 3, 2020). The survey was administered again between May 11 and 13, 2020, with a new sample of 502 adults to assess changes in proportion of U.S. adults wearing face coverings in public, as well as factors associated with use of face coverings in public. Use of a face covering was assessed by the following question: “In the past week, when you have gone outside of your home for work, grocery shopping, or other activities that involved interacting with other people, how often did you wear a cloth face covering that covered your nose and mouth?”. The primary outcome was any use of a face covering, defined as always, often, or sometimes using a face covering while in public.
Summary of Main Findings
Within one month, the proportion of respondents reporting wearing a cloth face covering when leaving their homes increased by 14.5%, from 61.9% in April to 76.4% in May. High variability in use of face coverings in the baseline survey was observed across age, race/ethnicity, income, employment, and census region; however, many of these differences leveled off by May, following significant increases in reported face coverings across demographic groups with lower reported use of face coverings in April. Behavioral constructs associated with face mask use included: positive attitudes towards use of face coverings, intentions to use face coverings in public, capacity to use a cloth face covering, perceived expectations to wear a face covering in public, and perceived effectiveness of cloth face coverings in preventing COVID-19 transmission.
The investigators derived constructs from well-established theoretical models of behavior change (e.g., Theory of Planned Behavior, Health Belief Model) to identify factors associated with cloth face coverings in public. Investigators also implemented quota sampling and applied analytic weights in order for survey responses to be representative of the U.S. adult population.
Despite recruitment and statistical techniques to make survey responses representative of the U.S. population, participants in an online survey may differ substantially from non-participants; the generalizability of these results is, therefore, questionable. As all survey responses were self-reported, results are susceptible to information and social desirability biases. While analyses controlled for specific demographic factors (i.e., age, race, census region), analyses were neither stratified across groups to identify stratum-specific factors associated with cloth face covering use. Critically, and most importantly, the authors did not did not conduct sub-analyses by frequency of face covering use (i.e., they did not disaggregate participants who always, often, and sometimes wore face coverings).
This is among the first studies to identify prevalence differences in, as well as theoretical behavioral constructs associated with, cloth face covering use in the U.S. population.
This review was posted on: 28 July 2020