Understanding the Touchscreen-based Nonvisual Target Acquisition Task Performance of Screen Reader Users
Hwayeon Joh, Yun Jung Lee, Uran Oh · 2022 · Proceedings of the 19th International Web for All Conference (W4A 2022) · doi:10.1145/3493612.3520454
Summary
This study investigates how screen reader users find and select targets on a smartphone touchscreen without visual cues, examining the effects of screen size, target size, and the ratio between them. Twelve participants with visual impairments (three totally blind, three with light perception only, one with low vision, and five with low vision from a national school for the blind in South Korea) completed two tasks on an iPhone 11: Task 1 involved finding a randomly placed single target by touch-exploring the screen, and Task 2 involved finding a specific labelled target among 16 candidates arranged in a 4x4 grid. The researchers systematically varied screen size (simulated via paper overlays at 400px, 200px, and 100px) and screen-to-target ratio (16:1 and 64:1), creating conditions that ranged from large screens with large targets to small screens with tiny targets (down to 25px, close to the W3C recommended minimum of 9mm). For Task 1, VoiceOver navigation was disabled to capture raw touch-explore traces; for Task 2, participants could freely use touch-to-explore or navigational swipes. Touch event coordinates were collected with timestamps, and sessions were video-recorded. The study collected 768 trials for Task 1 and 576 for Task 2, with outliers removed using the 1.5xIQR rule. Beyond performance metrics, the researchers performed a detailed trace analysis of 81 exploration traces from Task 1 to identify and classify the screen search strategies participants employed.
Key findings
Performance was significantly affected by both screen size and screen-to-target ratio. For Task 1, a one-way ANOVA revealed significant differences in completion time across conditions (F(3,33)=36.50, p<.001), with smaller screens and larger targets generally yielding faster performance and shorter travel distances. However, the screen-to-target ratio introduced a nuance: a very small target on a small screen (25px on 100px) was actually slower than a larger target on a bigger screen (100px on 400px), despite the smaller search space, because the target was too small relative to finger touch area. For Task 2 with distractors, medium screens with medium targets were significantly faster than other conditions (F(2,22)=20.25, p<.001). The trace analysis identified five distinct screen exploration strategies: zigzag (58%, the most common, with horizontal and diagonal variants), border-first (16%, scanning edges before center), pigtail (5%, continuous small circles), hybrid (14%, combining strategies), and other (7%). Participants tended to start searching from the top-left corner of the screen regardless of target location. Participants strongly preferred densely packed targets without gaps (10 of 12), and 6 of 12 preferred larger screens. The border-first strategy had the shortest travel distance, while scanning the middle area first was most efficient when target location was unknown. Participants used multiple strategies and switched between them when initial approaches failed.
Relevance
This research provides concrete, empirically grounded design guidance for mobile app developers and UI designers creating touchscreen interfaces for screen reader users. The key practical takeaway is that the screen-to-target ratio matters as much as absolute target size — designers should maximize the proportion of interactive screen area occupied by targets rather than simply meeting minimum size guidelines. The recommendation to use compact screen layouts with densely placed, gap-free targets challenges conventional design wisdom that emphasizes generous spacing between touch targets. For app icon grids, home screens, and grid-based navigation, this suggests that filling the available space with tappable targets (rather than leaving margins and gaps) actually improves non-visual usability. The taxonomy of five exploration strategies is valuable for anyone designing touch-based accessible experiences: knowing that most blind users explore in a zigzag pattern from the top-left corner means that the most important or frequently accessed targets should be placed in that region. The study also highlights the need for better integration between zoom and screen reader functionality on smartphones — currently, users must turn off the screen reader to zoom, which is a significant barrier.
Tags: touchscreen · screen readers · target acquisition · mobile accessibility · visual impairment · VoiceOver · interaction design · touch exploration