Measuring the Separability of Shape, Size, and Color in Scatterplots

Abstract

Figure 1: We examine how mark shape, size and color influence data interpretation in multiclass scatterplots. We replicate prior findings showing an interdependence between size and color and find evidence that shapes also bias size and color perceptions between different shape categories.

Reconfiguration Strategies with Composite Data Physicalizations

“…separability and integrality [27]) relates to proximity and atomic orientation changes in 3D space. Following 'data-agnostic' studies on 2D visual perception [3,11,30], we use data-agnostic 'abstract' 3D shapes to study interaction strategies based on proximity, orientation, and tangibility -not a specific dataset. This allows our study into interaction strategies to be independent of any dataset or domain, and more generally informative to future developments in this space.…”

Generative Design Inspiration for Glyphs with Diatoms

“…Difficulties associated with constructing glyphs also reflects the immensity of their design space. Beyond the initial selection of mark shapes and encoding channels, designers must be conscious of how the perception of mark shape, fill color, and mark size can interact [69], and they must weigh the effects of juxtaposing or superimposing marks on Gestalt perception within and between glyphs. Accordingly, there are varying approaches to glyph design, including experimenting with visual metaphors from nature or other media [48], appropriating semantically-related figurative elements and frames [15,39], and constructing taxonomies of abstract mark arrangements [51].…”

Which emphasis technique to use? Perception of emphasis techniques with varying distractors, backgrounds, and visualization types

“…Empirical studies of visual variables in visualization have generally evaluated their utility for design and their individual features for guiding visual perception. For example, Smart and Szafir 29 measure how size and shape affect color perception, Demiralp et al 30 developed the idea of ''perceptual kernels'' to measure the perceived similarities between pairs of visual variables (color, size, and shape), and Haroz and Whitney 31 explored the perceptual limits of attention for visual effects, investigating how visual feature type (color vs motion), layout, and variety of visual elements affect user performance, with findings suggesting that features that are known to be pre-attentive are adversely affected by arrangements and tasks that also require attention. Similar to both Haroz and Whitney and Demiralp et al, our experiments also address the question of whether a viewer can determine if a target is present on the screen; we expand on their findings by including a much larger number of effects, and explore a set of visualizations taken from the real world to increase our understanding of how layout and visual elements affect performance.…”