Using social media to determine specific causes of vaccine hesitancy in different regions could be a better way to create targeted messaging campaigns than traditional surveys, computer scientist Mayank Kejriwal, PhD, wrote in a Nov. 5 article for The Conversation.
Dr. Kejriwal said that while traditional survey methods can produce helpful data, detailed questions are expensive to administer, and sample sizes are often small because of cost constraints and nonresponsiveness.
He led research, accepted for publication in PLOS Digital Health, that uses geo-located tweets to predict the extent of vaccine hesitancy at the ZIP code level in U.S. metropolitan areas. The study found that by analyzing tweets, researchers could more accurately predict vaccine hesitancy by ZIP code than by using ZIP code attributes such as average home price and number of healthcare and social services facilities.
Using tweets, the research team developed algorithms that automatically collect potential causes of vaccine hesitancy in a given ZIP code, which public health officials could use to distribute targeted information and create strategic messaging to stop the spread of misinformation.