User interface (UI) funcitonality and aesthetics for color blind (CVD) users
Description
Color vision deficiency (CVD, color blindness) is the failure or decreased ability to distinguish between certain colors even under normal lighting conditions. There are over 300 million people worldwide with CVD including approx. 1 in 12 men (8\%) and 1 in 250 women (0.5\%). CVD can limit a user's ability to interact with websites and software packages that are otherwise basic commodities for the average user. User interface designers have taken various approaches to tackle the issue with some interfaces offering a high-contrast mode, and others integrating color-blind awareness into their website design process. The Web Content Accessibility Guidelines (WCAG) outline some best practices for maintaining accessibility that have been adopted and recommended by several governments; however, it is currently uncertain how this impacts perceived user functionality and if this could result in a reduced aesthetic look. In our work, we present a subjective user study to measure the loss of functionality and aesthetics as potentially seen by CVD observers for 20 popular websites and software packages. This dataset contains the results of the experiment (data_clean.csv) alongside the stimuli images.
Files
Steps to reproduce
To explore the impact on CVDs, we conducted a subjective user experiment. Specifically, we investigated the following two aspects: * Functionality: how well could a CVD user use the interface? I.e. in contrast to a non-CVD user, is usability reduced? Are key visual elements equally perceivable? * Aesthetics: according to classical aesthetics, is the clarity and orderliness of the UI retained when viewed by a CVD observer? In our experiment design, we address three types of color blindness: protanopia, deuteranopia, and tritanopia. As it would be challenging to recruit participants, especially with tritanopia, and a large sample would be required to establish the relative functionality and aesthetic scores by CVD vs. non-CVD users, we propose a full-reference simulation-based experiment, where participants without CVD can make a comparative judgment between the stimulus as seen by CVD and non-CVD users. Such a pipeline has the added benefit that it is suitable to be deployed in a commercial environment. To understand if there is a loss of functionality and/or aesthetics, we presented users with the original reference image as well as three distorted images that were identical in terms of content, showing what a CVD person might see. For content, we sampled 20 software in total across different industries (social, business, entertainment, music, shopping, food and beverage, travel, education, productivity, and SaaS). Within each sector, we picked the application or website based on general popularity in the UK, according to informal surveys. Participants viewed static screenshots. Specifically, in each software, we chose a common user journey that was relevant to the use of that software. To understand the functionality and aesthetics of high-contrast modes, we presented each application twice (1): distorted images from the reference image, (2): distorted images from high-contrast screenshots (but the reference left unaltered). In total, there were 120 stimuli (20 UIs x 3 screenshots x 2 modes of standard or high-contrast mode).
Institutions
- Perse School