Assessing Fit of Nontraditional Assistive Technologies
Adriane B. Randolph, Melody M. Moore Jackson · 2010 · ACM Transactions on Accessible Computing · doi:10.1145/1786774.1786777
Summary
This paper addresses a critical gap in assistive technology: how to systematically match users who have severe motor disabilities with nontraditional assistive technologies (NTATs) that use physiological signals rather than physical movement for computer control. The authors focus on two NTAT types: functional near-infrared (fNIR) imaging, which detects brain activity by measuring oxygenated blood volume through the skull, and galvanic skin response (GSR), which measures skin conductivity changes driven by the autonomic nervous system. These technologies are vital for people with conditions like ALS or locked-in syndrome who cannot use traditional input devices like keyboards, mice, or even eye-tracking. The researchers developed an Individual-Technology Fit (ITF) framework that profiles 28 user characteristics across demographic, physiological, and experience categories. They tested 38 participants—33 nondisabled individuals and 5 people with ALS—using both fNIR and GSR technologies. The BioGauges methodology measured each participant's ability to control their physiological signals to hit targets on screen. This exploratory design research approach aimed to identify which individual characteristics predict success with each technology, enabling practitioners to recommend the most appropriate NTAT for a given user.
Key findings
The study identified distinct predictor profiles for each technology. For fNIR, three characteristics significantly predicted control success: age (younger users performed better), caffeine consumption (non-consumers performed better), and years of education (more education correlated with better performance). For GSR, twelve characteristics were significant, condensing to six key variables: sex (men outperformed women), age (younger better), hair/skin color and texture (lighter features correlated with better control, possibly due to sensor-skin interactions), alcohol consumption (regular drinkers performed better, attributed to thinner blood and better circulation), meditation experience (non-meditators performed better, contrary to expectations), and video game experience (players of FPS, strategy, RPG, and simulation games performed better). Crucially, physical characteristics like paralysis and athleticism did NOT significantly affect fNIR control. This is encouraging news for people with ALS—as their condition progresses and motor function declines, they should still be able to control an fNIR device effectively. Age was the only shared significant predictor between technologies, with younger people performing better on both. The study found no significant performance differences between participants with ALS and nondisabled participants, supporting the validity of using nondisabled participants in exploratory research.
Relevance
This research provides a practical framework for assistive technology practitioners to match users with appropriate brain-computer interfaces. The NTAT checklist (Table VII) offers a simple assessment tool: seven questions about age, sex, caffeine use, education, meditation, video game experience, and alcohol consumption, plus observation of hair color, can guide technology recommendations. For fNIR, an ideal candidate is a younger adult with college education who doesn't regularly consume caffeine. For GSR, an ideal candidate is a younger adult male with lighter hair who doesn't meditate, plays video games, and drinks alcohol regularly. The finding that motor ability doesn't predict fNIR success is particularly valuable—it means people can continue using this technology as conditions like ALS progress. However, limitations include the small sample of disabled participants (only 5) and the laboratory setting. Future work should validate these findings with larger disabled populations and examine long-term training effects, as practice may eventually override some of these individual characteristic predictors.
Tags: brain-computer interface · assistive technology · motor disabilities · ALS · locked-in syndrome · user profiling · fNIR · GSR