Study objectives: To develop a sleepiness scale devoid of semantic or geometric elements.
Design: Subjects were asked to rank in order 7 cartoon faces representing degrees of sleepiness. We used Thurstone's scaling procedure to transform these rankings into an interval scale, which allowed us to eliminate 2 of the faces. The remaining 5 faces were ranked again using other subjects. In a validation study, subjects rated their perceived level of sleepiness using our scale and other sleepiness scales. Employed shiftworkers and school-going children used our scale to assess its practical applicability.
Settings: Research and diagnostic sleep laboratories, pre-primary to tertiary institutions, shift-working industry.
Participants: Ethnically diverse healthy and sleep-disordered adults (n = 490), and school-going children (n = 345).
Measurements and results: Our faces scale correlated with the Karolinska Sleepiness Scale (P < .05), the Stanford Sleepiness Scale (P < .04), and a visual analog scale measuring sleepiness (P < .0001). Shiftworkers showed a time-on-task effect on the evening shift (P < .0001) and a peak in sleepiness at 4:00 and 5:00 (P < .0001) on the night shift. Eight to 10 year old children appeared sleepier than older children throughout a school day (P < or = .02) and became sleepier as the day progressed (P < .0001). We confirmed that our scale measures sleepiness, uncontaminated by pain, anger, or happiness.
Conclusions: We have devised a sleepiness scale suitable for people too young or insufficiently educated to employ more-conventional scales. We envisage the scale being used for diagnostic, therapeutic, and research purposes.