Sociotechnical Considerations for Accessible Visualization Design

Abstract

Figure 1: As part of an inclusive design workshop at the Perkins School for the Blind, we created a 3D printed tactile translation of a time-series chart by William Playfair. In this paper, we show how these one-to-one translations, while based on existing best-practice guidelines for tactile graphics, can be pedagogically ineffective and incur prohibitive costs. ABSTRACTAccessibility-the process of designing for people with disabilities (PWD)-is an important but under-explored challenge in the visualization research community. Without careful attention, and if PWD are not included as equal participants throughout the process, there is a danger of perpetuating a vision-first approach to accessible design that marginalizes the lived experience of disability (e.g., by creating overly simplistic "sensory translations" that map visual to non-visual modalities in a one-to-one fashion). In this paper, we present a set of sociotechnical considerations for research in accessible visualization design, drawing on literature in disability studies, tactile information systems, and participatory methods. We identify that using state-of-the-art technologies may introduce more barriers to access than they remove, and that expectations of research novelty may not produce outcomes well-aligned with the needs of disability communities. Instead, to promote a more inclusive design process, we emphasize the importance of clearly communicating goals, following existing accessibility guidelines, and treating PWD as equal participants who are compensated for their specialized skills. To illustrate how these considerations can be applied in practice, we discuss a case study of an inclusive design workshop held in collaboration with the Perkins School for the Blind.

Accessible Visualization via Natural Language Descriptions: A Four-Level Model of Semantic Content

“…Facebook's automatic image descriptions, for example, have been characterized by technology educator Chancey Fleet as "famously useless in the Blind community" while "garner[ing] a ton of glowing reviews from mainstream outlets without being of much use to disabled people" [33,40]. Such concerns might be mitigated by developing and evaluating automatic methods with disabled readers, through participatory design processes [67].…”

“…While automatic methods researchers often note accessibility as a worthy motivation [27,30,31,78,83,84], evidently few have collaborated directly with disabled people [25,71] or consulted existing accessibility guidelines [67]. Doing so is more common to HCI and accessible media literatures [73,91], which broadly separate into two categories corresponding to the relative expertise of the description authors: those authored by experts (e.g., publishers of accessible media) and those authored by non-experts (e.g., via crowdsourcing or online platforms).…”

“…First, participants recruited through Prolific are usually not accustomed to completing lengthy tasks -our prior surveys and pilots suggested that these participants might contribute low-quality responses on "clickthrough" tasks if the task duration exceeded 15-20 minutes -and thus we asked each participant to rank only 5 of the 15 visualizations at a time. Second, given the difficulty of recruiting blind readers proficient with screen readers, we asked each blind participant to rank all 15 visualizations, and compensated them at a rate commensurate with their difficult-to-find expertise [67]. In this way, we recruited sufficient numbers of readers to ensure that each of the 15 visualization ranking tasks would be completed by 30 participants from both reader groups.…”

Designing equitable and inclusive visualizations: An underexplored facet of best practices for research and publishing

“…A third dimension of inclusivity concerns readers' physical dis/abilities and differences. Several ways in which dis/abilities affect readers' interactions with visualizations have been documented (Lee et al, 2020; Lundgard et al, 2019). For example, color vision deficiency (CVD) or color blindness can affect readers' ability to interpret visualizations that rely on color to convey information (Relvas, 2018).…”

“…For best practices when using alt‐text where possible, see Gustafsdottir (2021) and Sollinger (2014). It is worth noting that the visualization community is also addressing accessibility and inclusivity for other challenges, such as blindness (Lee et al, 2020; Lundgard et al, 2019), but those considerations are largely outside the scope of the decision‐making of individual engineering education authors and are not limited to visualizations, which are the focus of this editorial.…”