Download excel file with all results here: gapm.io/c19surv1
Read about the method below»
Free License
This data is free to use under Creative Common Attribution License BY 4.0 (License tex: https://gapm.io/cc_surveys)
“Which means: Please use it and redistribute it in any way you want, as long as you mention
“Free survey results from gapminder.org.””
Survey background
Gapminder is an independent educational non-profit specialized in global misconceptions. In March 2020 we developed a questionnaire with 10 questions about the Corona-pandemic. The questions were asked to people in five countries in March, using Google Surveys, and then we adjusted the set of questions sligthly and did a second wave of surveys in ten countries in May.
The study was managed by Maike Winters for Gapminder’s misconception project.
If you got questions please email: [email protected]
Method
We conducted two rounds of nationally representative surveys, targeting five countries in the first round (March 14-19, 2020) and 10 countries in the second round (May 18-23, 2020). A total of 2608 participants completed the first survey in the USA, the UK, Italy, Brazil and Japan. We excluded 551 participants (21%) from the analyses due to incomplete demographic data. In the second survey, 3695 respondents completed the surveys in the same countries as in the first survey, plus Australia, Canada, Germany, Mexico and the Netherlands. We excluded 117 (3%) respondents from the analysis in round two due to missing data – the share is lower in the second round because we only used the Google Opinion Reward App, whereby demographic data is available for the large majority of the respondents. Basic demographics of our samples are summarized to the right
This survey was administered to around 500 respondents in each of the following five countries: the UK, the USA, Italy, Brazil and Japan, between March 14-19, 2020. At that time, all of these countries were affected by the COVID-19 pandemic. Italy had the most confirmed COVID-19 cases per capita at the time of this survey and was in nationwide lockdown (ref OWID), while the number of confirmed cases was just beginning to rise in the other four countries. To be able to understand whether beliefs in misinformation changed over time, we developed a second questionnaire. This included some of the same questions as the first survey, but also included questions on willingness to take a future COVID-19 vaccine and levels of worry about contracting COVID-19. The overlapping questions concerned the origin of the coronavirus, the mortality of COVID-19 in comparison with the seasonal flu as well as disease prevention behaviors and distrust in information sources. The same five countries were surveyed in the second round (with also around 500 respondents in each country), as were 200 respondents in each of the following five countries: Australia, Canada, Germany, Mexico and the Netherlands. The second survey took place between May 18-23, 2020.
Google Surveys was used to disseminate the surveys, an online research platform designed to carry out nationwide surveys using ‘survey walls’ and the Google Opinion Reward App on Android. Respondents answer the questions anonymously, only an age range (with 10-year intervals, minimum age 18 years) sex and region are known. In the first round of surveys, a combination of the survey walls and the Google Opinion Reward App was employed. Because of the sensitive nature of asking about willingness to take a vaccine, the surveys were only targeted to Google Opinion Reward App in the second round of surveys. This is an opt-in only method, starting with a screening question to ensure that respondents know that questions about their health would be asked. To make the surveys as representative of each targeted country as possible, Google Surveys calculates survey weights per country based on a ranking method, whereby the weights are based on sex, age and region. Because our aim was to pool all surveys of the various countries, we replicated the raking method, and included country as an additional variable in the weight calculation.
About Google Consumer Surveys
See: Comparing Google Consumer Surveys to Existing Probability and Non-Probability Based Internet Surveys