If you see a survey pop up on your Facebook feed asking you about COVID-19, consider filling it out.
That information will help researchers at Carnegie Mellon University and other collaborating universities forecast the spread of the pandemic. Whether you have symptoms or not, they want to know. They don’t want your name, they just want to gather data.
Lots of data.
“We are trying to measure something for which there is no ground truth yet,” says Ryan Tibshirani, associate professor of statistics and machine learning at CMU.
No other source is available to verify the survey findings, he notes.
“So the only way for us to feel confident about the results is to get as many replies as possible,” notes Tibshirani, “That’s why we need to blast this out to as many people as possible.”
Partnering with Facebook, and their millions upon millions of users, is the best possible way to get the word out.
“We don’t have good data at this point regarding symptomatic infections,” says Tibshirani, who is also co-leader of CMU’s Delphi Research Group, one of two Influenza Forecasting Centers of Excellence designated last year by the Centers for Disease Control and Prevention (CDC). “People have been discouraged from visiting physician offices and hospitals. The only way to get this is with the survey.”
They’re seeking to measure data about symptoms at the county level. A rise in symptoms could easily indicate a coming spike in hospitalizations.
“The goal is to try to get a good idea of where COVID-19-like illnesses are now, and where it might be appearing next,” says Logan Brooks, a postdoctoral researcher with the Delphi Research Group. “To get a detailed picture nationally. For example, we want to know in certain counties … how many COVID-19-like cases there are, and how prevalent it is. Whether it’s going up or going down and how quickly. And then use that in combination with statistical and machine learning methods to get a good picture of where the pandemic is moving.”
The long-term goal of the Delphi Research Group is “to make epidemiological forecasting as universally accepted and useful as weather forecasting is today.”
Sick or not, they want to know.
“Even people who are healthy,” says Brooks. “We want healthy and sick people to fill out the survey. We want to know how common the sickness is. Both types of responses can improve the quality of the picture.”
The problem with getting information about COVID-19 transmission is that we really only know about the people who have gotten tested. There are lots of others out there, just staying home.
Those people are crucial in finding out where the virus is, and where it may be headed.
“It’s really up-to-date Internet data,” says Brooks. “You can get many users, many people who aren’t showing up in doctor’s offices and hospitals yet. It may provide advance notice, and a more detailed geographic picture.”
Facebook is providing users, but isn’t involved in the survey, notes Tibshirani. The CMU team is also working on getting Google, another of the world’s biggest tech firms, involved in the project.
The information collected will be used and shared as broadly as possible.
“We hope to be able to provide that to the states, to the public,” says Brooks. “But also more detailed information might be useful to other researchers in public health. They may investigate other types of questions, and improve our understanding of it.”