Age-specific predictive models of human performance

Abstract

Designers often struggle to create interfaces that are optimal for both younger and older adults, as they may interact differently with the same interface. Human-performance models have been used to aid designers in evaluating the efficiency of user interfaces. Can we create age-specific models to help designers create interfaces that are efficient for all age groups? We modeled a target acquisition task using published younger and older person parameters. While the younger model's mean prediction matches younger human data well (within 3.2%), the older model overestimates older users' mean task times by 34.6%. Further work should explore the influence of device type and the role of error-avoidance on parameter values for models of older adult interactions with technology.