Fig 1.
Seven-day rolling average of (a) COVID-19 tweet sentiment and (b) COVID-19 incident cases.
Table 1.
Statistical values for overall correlation.
Fig 2.
(a) Daily correlation coefficient between sentiment measures and daily new cases per 100,000 in each county and (b) the corresponding p-values and (c) z-scores.
Fig 3.
(a) Correlation coefficients between TBCOV sentiment and COVID-19 cases for all counties. (b) Anselin local Moran’s I cluster and outlier analysis of TBCOV correlation coefficients. An interactive version of this map can be found at
https://mygmu.maps.arcgis.com/apps/mapviewer/index.html?webmap=e73b409615824f9f9e019623b42ae664.
Fig 4.
Fig 4. TBCOV correlation coefficients visualized spatially for 3 different time periods. An interactive version of this map can be found at https://mygmu.maps.arcgis.com/apps/mapviewer/index.html?webmap=bc1qs8x6caxpkns9szm37d8csd6ej4nr67qx3hk732.
Table 2.
Regression results for different time periods.