David Broniatowski’s research investigates decision making under risk, big data and health, systems architecture, and computational analysis of group decision-making. What all of these research areas have in common is a drive to understand how information flows within technical and social systems, impacting decision-making and design.
“A lot of our models are built on eighteenth century assumptions about how people make decisions,” he says. “I work with experimental psychologists to understand how people actually make decisions.”
For example, Dr. Broniatowski recently co-authored a study, which found that patients are likely to expect antibiotics for viral infections, even though they are aware that antibiotics probably will not cure them. They do this because they compare the status quo—being sick— to the very unlikely possibility of getting better from antibiotics, while assuming that side effects of antibiotics are negligible.
In a related study, Dr. Broniatowski determined that people’s perceptions about vaccines also will drive their vaccination behaviors, and that these perceptions can be tracked on social media. In fact, the National Institutes of General Medical Sciences in the National Institutes of Health recently awarded Dr. Broniatowski a $1.5 million R01 grant to continue his study of how people use social media to share their rationales for vaccinating or refusing to vaccinate.
His work on social media builds on previous successful research that used Twitter to track the flu. Although several researchers are trying to use social media to track the flu, Dr. Broniatowski and his colleagues at Johns Hopkins University have had more success than others. Their algorithm can tell the difference, on both the local and national levels, between tweets indicating the person actually has the flu and those expressing only awareness of flu. Their technique was 93 percent accurate when compared to actual national influenza-like-illness data collected by the Centers for Disease Control and Prevention (CDC) during multiple flu seasons.
“Because the CDC’s data collection from hospitals and physicians involves a time lag, a system that uses Twitter might be able to reveal a spike in flu cases more quickly,” explains Dr. Broniatowski. “Data collected from Twitter may allow cities to do better surge planning for the flu or other epidemics because it can be gathered in real time.”
The scientific community has recognized the success of his technique through a letter in the prestigious journal Science. But Dr. Broniatowski also works to keep his research relevant to the public, and his studies have received a good deal of media attention, including reports in the Washington Post, National Journal, Men’s Health, and other media outlets.