Data visualization that "fits": Designing effective dashboards for healthcare providers, patients, and family caregivers to patients with diabetes

Abstract

Information dashboards are at-a-glance displays that are intended to highlight the most important information in order to help individuals solve problems and make better decisions using data. User-centered design of dashboards is necessary because different users have unique objectives when interacting with these displays. Due to advances in technology access to data and visualization tools it is now possible and creation of dashboards is now easier than ever. This also led to the popularity of infographics or the visualization that uses icons and metaphors to represent information on dashboards. Cognitive fit theory was used to explain the importance of considering the type of task in explaining the effectiveness of different visualization types (tables and graphs) and the type of information (spatial and symbolic) each visualization type emphasizes (Vessey, 1991). However, previous literature on visualization only focused on tables and graphs. While studies identified where and when graphs and tables were effective, research on infographics is still in its infancy. This study examined the effects of visualization type (table, graph, infographic) on accuracy, performance, and preference for two different task types (symbolic, spatial) and information levels (daily, monthly). Participants were 27 healthcare providers and 28 patients or family caregivers. Results indicated that tables and infographics were the best displays for symbolic tasks regardless of information type. Graphs were the best displays for monthly information and spatial tasks. Infographics were the most preferred displays overall. Consistent with previous studies, results indicated a mismatch between users' performance and preference.