As the World Health Organization (WHO), along with other organizations like the CDC and medical professionals, urges the world to limit physical contact to help slow the spread of COVID-19 many (businesses are allowing or directing their employees) to use strategies like working from home to promote social distancing efforts. In some areas, ‘shelter in place’ ordinances that allow for only essential travel outside the home are forcing workers (whose job permits) to resort to teleworking. Additionally, schools across the country are shutting down and going digital.
As the Novel Coronavirus continues to spread around the world, Premise has been working to monitor and understand people’s sentiment toward COVID-19 and understand how their lives are changing as a result. You can check out all of the Global Impact Survey data we have been collecting here.
For this COVID-19 snapshot, we will focus on the United States. With the help of our Data Science team, we were able to determine if there were statistically significant differences in attitudes between respondents in states with the highest number of COVID-19 cases (Group A) and respondents in states with the lowest number of COVID-19 cases (Group B) on topics related to coronavirus.
After looking at how COVID-19 has impacted American respondent’s willingness to visit public spaces, we are looking at another similar topic – working from home. To do that we asked the question, “Have you or a member of your household decided to stay home from work or school because of the Coronavirus (COVID-19)?”
% of people in Group A that said “Yes” = 32.06%
% of people in Group B that said “Yes = 23.15%
What we found is that there is a statistically significant difference between the two proportions (p-value < 0.01), suggesting that the percentage of people in Group A staying at home is higher than that of people in Group B. This is consistent with the statistical finding with the data on people visiting public spaces. As cases rise across the United States and local governments begin to take more extensive measures to protect their populations we will see if these numbers change.
For each question, we obtained the proportions of Groups A and B. Then, we conducted a 2-sample z-test (a statistical hypothesis test) in Python and calculated a p-value to assess whether there was a statistically significant difference between the two groups’ proportions.
Group A was comprised of respondents of the three states with the highest number of COVID-19 cases, New York, Washington and California (according to CNN on March 17th, 2020). Group A totaled 1,883 observations,*
Group B was comprised of respondents from the 14 US states with the lowest number of COVID-19 cases according to CNN on March 17th, 2020—West Virginia, North Dakota, Alaska, Wyoming, Missouri, Idaho, Montana, Hawaii, Delaware, Vermont, Oklahoma, South Dakota, Kansas and Mississippi. We chose to use 14 states in order to not have a huge disparity between the two sample sizes of the two groups. Group B totaled 1,151 observations.*
*Total Observations for the United States: 12,861 [As of 12:13 PM on Tuesday, March 17, 2020]