Data Visualization and Reporting
Robert Blagg, PhD
Scientist
UCLA, United States
Location: Room 104
Abstract Information: Research indicates that attitudes, processes, and other factors influence the use of evaluative information in policy and practice decisions (Palinkas et al., 2016). Furthermore, individuals’ approaches to obtaining and using evaluative information provide practical insights on how agencies may address barriers between evaluation and practice (Parrish & Rubin, 2011). Data visualization provides a dynamic way to meet evaluation audiences’ needs and uses of evaluative information. Successfully engaging audiences with data visualization requires carefully considering intended uses of intended users (Patton, 2011), while incorporating strong design principles that leverage visual perception. To ensure data visualization solutions are useful and sustainable, the evaluator can build the capacity of audiences to consume and disseminate visualization, integrate visual feedback elements, and position visualization as a key to the evaluation information feedback loop. In this session, attendees will see demonstration of utilization focused data visualization principles, including how users perceive data visually (e.g., Kosara, 2016), and the elements of data visualization that have been found to support utilization and the engagement of audiences (e.g., Patton, 2008). Key aspects of designing visualizations for user action will be highlighted, including: 1) collecting and preparing survey data for visualization, 2) designing visualizations that respect visual perception, and 3) curating visualizations to focus users on actionable findings that engender ownership of data.
Relevance Statement: Central to evaluation as a profession, and specifically crucial to ensuring the utilization of evaluative information, is the translation of complex ideas to diverse stakeholders at multiple points throughout an evaluation (e.g., Azzam & Evergreen, 2013). Stakeholders to evaluation may vary greatly in the extent of their interest, familiarity, or expertise with data regarding an evaluand. Well-designed data visualization can provide a platform for engaging stakeholders in ways that are tailored to their interest, familiarity, and expertise. However, research indicates that visualizations are not always designed effectively enough to reliably spur users to act. As illustration, Evergreen (2011) found that evaluators rated graphics as the least utilized areas of reports. Kosslyn et al. (2012) found frequent misuse of communication principles in visualizations, most notably making them overly complicated for the message intended to be conveyed. But other research indicates that this does not need to be the case. Specifically, visualizations that are more thoughtful and engaging demand stakeholders’ attention and can facilitate comprehension of findings (Hillstrom & Yantis, 1994; Treisman, 1988). For example, Cawthon and Moere (2007) found that information consumers were more willing to spend additional time absorbing and interpreting data to the extent visuals were well designed. As such data visualizations can encourage stakeholders to engage with an evaluation on their own terms, curated by evaluators, by asking and answering their own often unique questions. Modern data visualization tools and platforms (e.g., Tableau, D3, Microsoft BI Desktop, SAS, and a growing list of others) allow stakeholders to share (e.g., online, via email, text message, paper, etc.) visualizations tailored to the evaluation findings they determine to be most relevant to the circle of stakeholders they wish to engage. Designing evaluations to integrate such a level of engagement, at multiple stages, may serve to encourage use (e.g., Patton 2008), by supporting engagement, good design, and capacity building and sustainability, support greater understanding and discussion that can bring meaning and relevance to evaluation data and make user action more likely.