From extraneous noise to categorizable signatures: Using multi-scale analyses to assess implicit interaction needs of older adults with visual impairments

The holistic understanding of human-computer interaction (HCI) is increasingly important, especially given the impending influx of older users who present dynamic needs that evolve with age. This study explores pupillary response behavior (PRB) during computer interaction to identify underlying differences between older adults of varying ocular profiles. PRB was measured from two groups of individuals diagnosed with Age-related Macular Degeneration (AMD) and a visually healthy control group. Unconventional analytical techniques - wavelet-based multifractal analyses - were used to identify PRB anomalies resulting from the effects of aging and/or ocular pathology. A distribution of regularity indices was extracted from the data signals to reveal signatures of PRB change patterns. One characteristic of the multifractal spectrum, Left Slope (LS), fully distinguished the user groups, revealing trends of increasing PRB irregularity with increasing levels of ocular dysfunction.