COVID-19 Preventative Health Survey 2020-2021
Afghanistan, Algeria, Angola, Argentina, Australia, Azerbaijan, Bangladesh, Bolivia, Brazil, Cambodi, 2020 - 2021
Get MicrodataIdentification
WLD_2020_CPHS_v01_M
COVID-19 Preventative Health Survey 2020-2021
| Name | Country code |
|---|---|
| Afghanistan | AFG |
| Algeria | DZA |
| Angola | AGO |
| Argentina | ARG |
| Australia | AUS |
| Azerbaijan | AZE |
| Bangladesh | BGD |
| Bolivia | BOL |
| Brazil | BRA |
| Cambodia | KHM |
| Cameroon | CMR |
| Canada | CAN |
| Chile | CHL |
| Colombia | COL |
| Cote d'Ivoire | CIV |
| Ecuador | ECU |
| Egypt, Arab Rep. | EGY |
| Estonia | EST |
| France | FRA |
| Georgia | GEO |
| Germany | DEU |
| Ghana | GHA |
| Guatemala | GTM |
| Honduras | HND |
| India | IND |
| Indonesia | IDN |
| Italy | ITA |
| Iraq | IRQ |
| Japan | JPN |
| Jamaica | JAM |
| Kazakhstan | KAZ |
| Kenya | KEN |
| Korea, Rep. | KOR |
| Malaysia | MYS |
| Mexico | MEX |
| Mongolia | MNG |
| Morocco | MAR |
| Mozambique | MOZ |
| Myanmar | MMR |
| Nepal | NPL |
| Netherlands | NLD |
| Nigeria | NGA |
| Pakistan | PAK |
| Peru | PER |
| Philippines | PHL |
| Poland | POL |
| Portugal | PRT |
| Romania | ROU |
| Senegal | SEN |
| Singapore | SGP |
| South Africa | ZAF |
| Spain | ESP |
| Sri Lanka | LKA |
| Sudan | SDN |
| Taiwan, China | TWA |
| Tanzania | TZA |
| Thailand | THA |
| Trinidad and Tobago | TTO |
| Turkey | TUR |
| Uganda | UGA |
| Ukraine | UKR |
| United Arab Emirates | ARE |
| United Kingdom | GBR |
| United States | USA |
| Uruguay | URY |
| Venezuela, RB | VEN |
| Vietnam | VNM |
Demographic and Health Survey [hh/dhs]
The COVID-19 Preventive Health Survey was designed to help policymakers and health researchers better monitor and understand people’s knowledge, attitudes and practices about COVID-19 to improve communications and their response to the pandemic. This survey is conducted in partnership between Facebook, the Massachusetts Institute of Technology (MIT), and Johns Hopkins University (JHU), with advice from the World Health Organization. Sampled users see the invitation at the top of their News Feed, but the surveys are collected off the Facebook app and the Facebook company does not collect or receive individual survey responses. The survey asks users to self-report their adherence to preventive measures, such as wearing masks and what they know about COVID-19, including symptoms of the disease, risk factors and how their community is handling the pandemic.
Sample survey data [ssd]
- Public Aggregate Data: Subnational levels
- Microdata through Facebook Data for Good program: Individual level
Scope
This survey was conducted off-platform and covers self-reported preventive behaviors and knowledge and attitudes towards COVID-19 vaccines.
Coverage
This survey was fielded in 67 countries and territories.
Wave Countries and Territories: Argentina, Bangladesh, Brazil, Colombia, Egypt, France, Germany, India, Indonesia, Italy, Japan, Malaysia, Mexico, Nigeria, Pakistan, Philippines, Poland, Romania, Thailand, Turkey, United Kingdom, United States, Vietnam
Snapshot Countries and Territories: Afghanistan, Algeria, Angola, Australia, Azerbaijan, Bolivia, Cambodia, Cameroon, Canada, Chile, Cote d’Ivoire, Ecuador, Estonia, Georgia, Ghana, Guatemala, Honduras, Iraq, Jamaica, Kazakhstan, Kenya, Mongolia, Morocco, Mozambique, Myanmar, Nepal, Netherlands, Peru, Portugal, Senegal, Singapore, South Africa, South Korea, Spain, Sri Lanka, Sudan, Taiwan, Tanzania, Trinidad & Tobago, Uganda, Ukraine, United Arab Emirates, Uruguay, Venezuela
The target population consists of active Facebook users. The sampling frame is active Facebook users with ages 18+, which includes users living within 23 countries or territories. The sampling frame is restricted to people who use Facebook in one of the supported locales.
Producers and sponsors
| Name |
|---|
| Facebook Data for Good |
| Massachusetts Institute of Technology (MIT) |
| Johns Hopkins University (JHU) |
Sampling
The target population consists of active Facebook users. The sampling frame is active Facebook users with ages 18+, which includes users living within 23 countries or territories. The sampling frame is restricted to people who use Facebook in one of the supported locales.
The Facebook app invites a sample of adult users to take an optional, off-Facebook survey through an invitation at the top of their Facebook News Feed. Users who click on the invitation are redirected to a Qualtrics page hosted by MIT where they are informed about the survey and can take the survey. While MIT designs, collects, and analyzes the survey data, Facebook provides assistance with questionnaire translation, survey sampling and recruitment, and statistical bias correction.
Response rates to online surveys vary widely depending on a number of factors including survey length, region, strength of the relationship with invitees, incentive mechanisms, invite copy, interest of respondents in the topic and survey design.
Survey instrument
The survey includes questions about self-reported preventive behaviors and knowledge and attitudes towards COVID-19 vaccines. The survey instrument is managed by MIT and available in more than 55 languages. Two versions of the survey were fielded across 67 countries and territories. Countries with sufficient sample sizes receive a “Wave Survey” that is fielded every 2 weeks between July 2020 and March 2021. The rest of the countries receive a periodic “Snapshot Survey”. Snapshot and wave surveys were developed based on feedback from global health partners so that information could be collected that is helpful to inform public health responses even in areas with fewer survey respondents. As of Spring 2021, some questions from the survey have been merged with the larger Covid-19 Trends and Impact Survey. The full survey instrument is available here.
The snapshot survey was fielded to 44 countries and territories with a one-time sample over a 2 week period. A follow up sample was done in late 2020 of snapshot countries and territories to provide updated information.
The wave survey was fielded to 23 countries and territories with repeated, bi-monthly cross-sections. Each of the 8 waves is two weeks long. Sampled users may be invited to take the survey again in subsequent weeks, depending on the density of their area. However, the responses of sampled users who participate more than once will not be linked longitudinally.
Data collection
| Start | End |
|---|---|
| 2020-07 | 2021-03 |
Facebook users are invited to take an off-platform survey conducted by MIT to self-report their adherence to preventive measures, such as washing hands and wearing masks, and what they know about COVID-19, including symptoms of the disease, risk factors and how their community is handling the pandemic. Facebook does not host the survey or collect survey responses, and only has access to public, aggregated survey data provided by MIT. To ensure that the survey sample more accurately reflects the characteristics of the population represented in the data, Facebook shares a random ID number and a single statistic, known as a weight value, that doesn’t identify a survey respondent but helps correct for survey sample bias. Facebook also shares the survey respondent’s language preference with MIT. Facebook doesn’t share who took the survey with MIT, and MIT doesn’t share individual survey responses with us.
Data appraisal
Any survey data is prone to several forms of error and biases that need to be considered to understand how closely the results reflect the intended population. In particular, the following components of the total survey error are noteworthy:
Sampling error is a natural characteristic of every survey based on samples and reflects the uncertainty in any survey result that is attributable to the fact that not the whole population is surveyed.
Facebook provides MIT (and other researchers) with analytic weights that adjust for non-response and coverage biases. Making adjustments using the weights ensures that the sample more accurately reflects the characteristics of the target population represented.
Non-Response Bias
This means that some sampled users are more likely to respond to the survey than others. To adjust for this, Facebook calculates the inverse probability that sampled users complete the survey using their self-reported age and gender as well as other characteristics known to correlate with nonresponse. Then these inverse probabilities are used to create weights for responses, after which the survey sample reflects the active adult user population on the Facebook app.
Coverage Bias
This means not everyone in every country has a Facebook app account or uses their account regularly. To adjust for this, Facebook adjusts the weights created in the first step even further so that the distribution of age, gender, and administrative region of residence in the survey sample reflects that of the general population.
Access policy
Facebook Data for Good.
Facebook - Data for Good.
https://dataforgood.facebook.com/dfg/tools/covid-19-preventative-health-survey
Depositor information
| Name | Affiliation |
|---|---|
| Joshua Seth Wimpey | World Bank |
Data Access
| Is signing of a confidentiality declaration required? |
|---|
| yes |
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Example:
Facebook Data for Good, Johns Hopkins University, Massachusetts Institute of Technology. World: COVID-19 Preventative Health Survey 2020-2021. Ref: WLD_2020_CPHS_v01_M. Downloaded from [url] on [date]
Metadata production
DDI_WLD_2020_CPHS_v01_M_WB
| Name | Affiliation | Role |
|---|---|---|
| Development Data Group | World Bank Group | Documentation of the DDI |
2021-10-27
Metadata version
Version 01 (October 2021)