Operational Fluency and Emotional Legibility: A Layered Account of Affective Self-Report by Older Adults

Compact affective interfaces must make reporting both easy and meaningful. We evaluated Feeling Tracker through a 21-day home deployment and a separate four-condition experiment, each with 58 older adults. At home, ease ratings increased from 5.53 to 6.02, while emotional legibility increased from 4.79 to 5.07 on seven-point scales. The gap persisted among frequent users and widened among intermittent users; frequent users improved on both items. Among 37 continuing users, distinct categories declined from 2.65 to 2.34 when three weekly selections were compared. In the experiment, spatially arranging the labels increased agreement with author-defined scenario targets from 0.63 to 0.71, and adding colour shortened modelled selection time by a further 11%. The full cue set supported higher target agreement and confidence and 28% shorter selection time than labels alone. Fixed faces added no clearly resolved further gain over the coloured grid. These findings identify complementary design priorities: spatial layout supports selection, colour aids orientation, and category boundaries require explicit semantic calibration. The study-derived provisional five-layer diagnostic framework translates these results into targeted design and evaluation questions for compact affective self-report.