A Micro-Phenomenological Lens for Evaluating Narrative Visualization.

Narrative visualizations engage audience in data stories, evoking emotions by using narrative patterns, rhetoric, visual design, and content among other strategies. How these elements combine to influence user experiences is complex and difficult to measure using empirical methods. This is partly due to the fact that narrative visualizations influence audiences affectively and implicitly [1]–[3]. Evaluations of narrative visualizations that aim to better understand these mechanisms should capture this rich complexity by focusing on gathering descriptions of lived experience. Micro-phenomenology, a rigorous set of methods developed for soliciting descriptions of experiences, has empirically been shown to improve recollection of otherwise implicit aspects of experience [4]. Building on work using micro-phenomenological interviews to evaluate static visualizations [5], we apply these methods to interactive narrative visualizations. We conducted a small study to explore the potential of these methods in this context. Our findings reveal how narrative patterns and designs influence affective states, how they support various forms of exploratory analysis, and how they can facilitate or hinder non-analytical reflection such as the imagining of stories described within visualizations. These types of insights can inform future designs and help researchers understand how techniques employed in narrative visualizations influence users in specific and often implicit ways.