Quantifying behavior change during the first year of the COVID-19 pandemic in the United States
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (ScreenIT)
Abstract
Background
During the first year of the COVID-19 pandemic, the most effective way to reduce transmission and to protect oneself was to reduce contact with others. However, it is unclear how behavior changed, despite numerous surveys about peoples’ attitudes and actions during the pandemic and public health efforts to influence behavior.
Methods
We used two sources of data to quantify changes in behavior at the county level during the first year of the pandemic in the United States: aggregated mobile device (smartphone) location data to approximate the fraction of people staying at home each day and digital invitation data to capture the number and size of social gatherings.
Results
Between mid-March to early April 2020, the number of events fell and the fraction of devices staying at home peaked, independently of when states issued emergency orders or stay-at-home recommendations. Activity began to recover in May or June, with later rebounds in counties that suffered an early spring wave of reported COVID-19 cases. Counties with high incidence in the summer had more events, higher mobility, and less stringent state-level COVID-related restrictions the month before than counties with low incidence. Counties with high incidence in early fall stayed at home less and had less stringent state-level COVID-related restrictions in October, when cases began to rise in some parts of the US. During the early months of the pandemic, the number of events was inversely correlated with the fraction of devices staying at home, but after the fall of 2020 mobility appeared to stay constant as the number of events fell. Greater changes in behavior were observed in counties where a larger fraction voted for Biden in the 2020 US Presidential election. The number of people invited per event dropped gradually throughout the first year of the pandemic.
Conclusions
The mobility and events datasets uncovered different kinds of behavioral responses to the pandemic. Our results indicate that people did in fact change their behavior in ways that likely reduced COVID exposure and transmission, though the degree of change appeared to be affected by political views. Though the mobility data captured the initial massive behavior changes in the first months of the pandemic, the digital invitation data, presented here for the first time, continued to show large changes in behavior later in the first year of the pandemic.
Article activity feed
-
SciScore for 10.1101/2022.01.10.22268799: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
NIH rigor criteria are not applicable to paper type.Table 2: Resources
No key resources detected.
Results from OddPub: Thank you for sharing your data.
Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:There were several limitations to our analysis. First, we assumed that Evites events represented in-person events and cancellation and RSVP data was reliable. Second, we are not able to disentangle the impacts of policy and other drivers of behavior change. Similarly, policy can be local, complicated, and heterogenous, making it difficult to compare nationally. Additionally, messaging from government outside of formal policies was not included …
SciScore for 10.1101/2022.01.10.22268799: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
NIH rigor criteria are not applicable to paper type.Table 2: Resources
No key resources detected.
Results from OddPub: Thank you for sharing your data.
Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:There were several limitations to our analysis. First, we assumed that Evites events represented in-person events and cancellation and RSVP data was reliable. Second, we are not able to disentangle the impacts of policy and other drivers of behavior change. Similarly, policy can be local, complicated, and heterogenous, making it difficult to compare nationally. Additionally, messaging from government outside of formal policies was not included here but may have influenced behavior [Pink et al 2021]. We acknowledge that factors we did not include in our analyses, such as weather and school closures, could have played a major role in SARS-CoV-2 transmission. Finally, we use ecological comparisons, and cannot compare individual-level behavior using these methods. We know that Evite users, those tracked by SafeGraph [Squire et al 2020, Grantz et al 2020, SafeGraph 2021], and voters [Verba et al 1995, Igielnik et al 2021] do not represent the general population. While Evites are most used in higher-income suburbs and major metropolitan areas, COVID incidence has been highest in rural and lower income counties [Li et al 2021]. However, the millions of events per year facilitated by Evite represent a unique data source for tracking trends in social gatherings. The first year of the pandemic enabled us to observe and quantify the relevant processes as well as to determine the appropriate geopolitical scales on which they act. Our analysis demonstrates the complexity of forces that dr...
Results from TrialIdentifier: No clinical trial numbers were referenced.
Results from Barzooka: We did not find any issues relating to the usage of bar graphs.
Results from JetFighter: We did not find any issues relating to colormaps.
Results from rtransparent:- Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
- Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
- No protocol registration statement was detected.
Results from scite Reference Check: We found no unreliable references.
-
