In a within-subjects experiment with 128 adults aged 65+ (mean age 71.5, 70% female), simpler instructions lacking salient visual cues proved more acceptable than visually enriched ones for complex functional cognitive games. Self-efficacy emerged as a key moderator: individuals with lower confidence in their cognitive abilities showed reduced receptivity to computerized cognitive training (CCT), except when the game was highly familiar. Higher cognitive self-confidence consistently predicted greater program acceptability.
This preprint, not yet peer-reviewed, challenges the intuitive design assumption that richer visual guidance uniformly improves technology onboarding for older users. The finding aligns with cognitive load theory — adding visual cues to already-demanding tasks may compete for the same limited attentional resources the training is meant to exercise, creating friction rather than clarity. For developers of brain-training platforms targeting aging populations, the implication is pointed: personalization matters more than embellishment.
Limitations are notable. The sample is relatively small, predominantly female, and confined to cognitively unimpaired adults, limiting generalizability to those with mild cognitive impairment — arguably the highest-need population. The within-subjects design also risks order and fatigue effects. Mechanistically, the study is observational and self-report-heavy, making causal claims about self-efficacy's role preliminary. Still, this work offers a practically actionable signal: cognitive training interfaces should default to simplicity and adapt to users' self-assessed confidence, not assume that more guidance is always better.
