What is it about?
We looked at accessibility problems in mobile apps from the point of view of the people building them. We collected 256 closed GitHub issues labelled as accessibility problems in open-source Android apps written in Kotlin, and read every one by hand. Two reviewers coded the issues independently and met to resolve disagreements, producing a taxonomy of 185 recurring problems in three areas: assistive technology support (135 issues), visual accessibility (32), and touch and gesture interaction (18). The most frequent barriers are missing or vague alternative text, decorative images announced as if they carried meaning, controls that cannot receive keyboard or switch focus, states such as checked or expanded that are never announced, labels that are not programmatically linked to their input field, low colour contrast, and touch targets smaller than the recommended size.
Featured Image
Photo by Airalo on Unsplash
Why is it important?
Most of what we know about mobile accessibility comes from automated scanners run over large sets of apps, or from studies of what end users complain about. Both miss the middle ground: what the people writing the code actually run into and fix. Reading real issue reports surfaced a clear pattern. Most barriers are not missing elements, they are elements with wrong or missing semantics. The button is there and looks correct on screen; it simply does not tell a screen reader what it is, what state it is in, or that it changed. That distinction matters because it is exactly the class of defect that current static and dynamic analysis tools handle worst: they reliably flag an image with no description, but they are weak at judging whether a description is meaningful, whether the focus order follows the visual layout, or whether a dynamic update was announced at all. The taxonomy gives tool builders a concrete list of under-covered checks, and gives development teams a checklist grounded in problems other developers have already hit and resolved.
Perspectives
This study is the starting point of my master's research on accessibility in mobile applications. What struck me while coding the issues was how often the failure was invisible both to the automated tools and to a sighted developer looking at the screen: the layout was correct, the control worked on tap, and nothing was announced to anyone using a screen reader. That is why we chose to mine issue trackers rather than run another scanner. We see this taxonomy as an initial, extensible artifact: we are now extending it to newer frameworks such as Jetpack Compose and Flutter, and studying how these issues get resolved over time.
Benjamin Gonzalez Briones
Pontificia Universidad Catolica de Chile
Read the Original
This page is a summary of: Characterizing Real-World Accessibility Issues Reported in Kotlin Mobile Apps, July 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3803437.3805568.
You can read the full text:
Resources
Contributors
The following have contributed to this page
