Exploring Methods to Improve Pen-Based Menu Selection for Younger and Older Adults
Karyn Moffatt, Joanna McGrenere · 2009 · ACM Transactions on Accessible Computing · doi:10.1145/1525840.1525843
Summary
This study investigates why users struggle with pen-based menu selection on tablet PCs, particularly older adults, and tests two approaches for reducing errors. Building on a baseline study that identified three error types (slipping, drifting, and "missing just below"), the authors focus on missing just below errors—when a user's tap distribution is shifted downward, causing frequent selection of the top edge of the item below the target. They developed two solutions: a "reassigned edge" approach where tapping the top edge selects the item above (shifting the motor target to match where users actually tap), and a "deactivated edge" approach where taps on the top edge are ignored, requiring the user to tap again. A controlled experiment with 24 participants (12 younger ages 19-30, 12 older ages 66-81) compared these approaches against traditional menus across 648 trials each. Participants completed standardized motor skill assessments including reaction time, Purdue Pegboard, digit symbol substitution, and hand steadiness tests.
Key findings
Only the deactivated edge approach showed overall promise, achieving higher net benefit than both traditional and reassigned menus. The reassigned edge prevented 68 missing-just-below errors but introduced 50 new errors on the target's top edge, negating its benefit. The most significant finding was the discovery of three distinct user groups through K-means cluster analysis: "low hitters" (7 participants) whose tap distribution shifted downward, "high hitters" (10 participants) whose taps shifted upward causing errors on the item above, and "neutrals" (7 participants) with no consistent bias. Critically, all but one older participant fell into an error-prone group, and 8 of 10 high hitters were older adults—a pattern that would not have been identified studying only younger users. For low hitters, both experimental approaches helped; for high hitters, the reassigned edge actually worsened performance. Older participants took significantly longer (2.4-2.5 seconds vs. ~1 second for younger) and required more taps with the deactivated condition, yet this did not significantly increase their selection time.
Relevance
This research challenges the dominant HCI focus on speed optimization, arguing that error reduction matters more for users who find errors frustrating and costly to recover from—particularly older adults. The discovery of opposing error patterns (low vs. high hitters) has important design implications: no single interface modification will work for all users, and systems may need to detect individual tap distributions before applying customized support. For accessibility practitioners, the study demonstrates why including older adults in research is essential—the high-hitter pattern would have been missed entirely with a younger-only sample. The findings also highlight the growing importance of pen-based accessibility as tablets and styluses are increasingly used in assistive technology applications. The subjective response data—showing users disliked the deactivated approach despite its performance benefits—underscores the importance of feedback design when modifying standard interaction patterns.
Tags: pen input · tablet PC · older adults · target acquisition · motor impairments · menu design · Fitts Law · input techniques · aging