Collaborative Research: Framework for Better Visual Communication Practices

Every day, people use scientific information to make decisions that affect their lives. Consequently, it is critical that the results of scientific research on topics impacting the public are communicated effectively. Data visualization is an important part of scientific communication, yet much guidance on the clear and effective design of data graphics has not been tested among a representative population of U.S. adults. The researchers examine how accurately users interpret the information conveyed in different types of data visualizations. The findings support evidence-based recommendations that aid in the fair and equitable communication of scientific information across demographic groups and subpopulations. This research advances understanding of data visualization as a method of communication and provide comprehensive data on viewer interpretation and understanding of data graphics among the adult U.S. population.

The researchers implement an online survey through a web portal to 2,000 respondents from a probability-based, nationally representative panel of U.S. adults. Respondents are shown multiple types of data visualizations with various design elements, such as bar and line charts with varying levels of supporting context and asked questions that measure their understanding and interpretation of the data presented. Survey questions then ask about basic interpretation of the values presented in a chart and open-ended responses about what conclusions respondents can draw about the data shown. The researchers analyze the data using generalized linear models, natural language processing, and text analysis techniques to determine the effect of different elements of data visualization design, such as the type of chart or supporting context, on user understanding, and the extent to which understanding may vary across demographic groups.