Skewed Dual Normal Distribution Model: Predicting 1D Touch Pointing Success Rate for Targets Near Screen Edges

A research team from Meiji University has published a groundbreaking paper, 'Skewed Dual Normal Distribution Model: Predicting 1D Touch Pointing Success Rate for Targets Near Screen Edges,' accepted at CHI 2026. The work directly addresses a major blind spot in UI design: typical predictive models for touch accuracy exclude targets near screen edges, yet design constraints and scrollable interfaces frequently place interactive elements there. The researchers propose a novel statistical model that accounts for how a nearby screen edge skews the distribution of user taps, fundamentally challenging established design principles.

The core finding, from two smartphone experiments, is that as a target approaches the screen edge, the peak of the tap distribution shifts toward it while the tail extends away. Most surprisingly, the success rate actually improves when the target physically touches the edge, suggesting users employ an effective 'tapping the target together with the edge' strategy. By modeling this skew, their framework can predict success rates across all screen positions, finally providing data-backed guidance for designing buttons, icons, and menus near edges. This enables the creation of next-generation UI design support tools with complete screen coverage, potentially improving usability in millions of mobile apps.