Modality and Uncertainty in Data Visualizations: A Corpus Approach to the Use of Connecting Lines

Abstract

In data visualizations, connecting lines may have various semiotic functions, including the semiotic potential of indicating modality and uncertainty. The goal of this article is to find out how this semiotic potential is realized in current best practices of data visualizations and what conventions exist for the visual manifestations of these functions. This issue is addressed by using a corpus-based approach and a two-level analysis method within a social semiotic framework. First, the article offers a theoretical discussion on how the concepts of modality and uncertainty interrelate. Second, a method for investigating how these concepts are visualized at different levels is presented. Third, a corpus analysis including 163 award-winning data visualizations is presented. The results indicate the existence of certain conventions for visual modality markers, and thus offer new insights relevant for both design theory and practice.

Number and quality of diagrams in scholarly publications is associated with number of citations

“…Our method follows Lee et al [18], adapted to use a manual extraction process in order to reduce systematic omission and make use of the validity ensuring method of Lechner [17]. paper they were extracted from.…”

“…5. Following the method of Lechner [17], inter-rater reliability was measured to validate scoring of guidelines compliance, on a subset of 15% of the resulting NN system diagram (17/119). This resulted in 204 pairs of pieces of metadata scored as "true", "false", or "not applicable".…”

“…Corpus analysis of diagrams is nascent, with recent analysis into connecting lines in data visualisations [17]. We use a corpus-based approach to examine diagrams within a wider social context, and have designed our approach to leverage existing document-level citation metrics, allowing quantitative analysis.…”