Author's response to letter to the editor [Refers to: Tomoyuki Kawada, Hiroko Suzuki, Takako Shimizu, Masao Katsumata Polygraphic sleep latency and subjective sleepiness by visual analog scale (VAS): The limitation of VAS Applied Ergonomics, Volume 43, Issue 1, January 2012, Pages 266]

Kawada et al. (2011) report a significant but moderately low correlation (r= -0.44) between subjective and objective sleepiness, measured by responses on a visual analogue scale (VAS) and sleep onset latency (SOL), respectively. Their correlation was calculated using single measures of VAS and SOL associated with a (daytime?) nap collected with 50 participants. The low correlation in this study was used to emphasize the point that VAS determined subjective sleepiness cannot be used as a substitute for objective sleepiness measured by SOL. Indeed, acceptable correlations for good measures of validity would normally have to be >0.70. Therefore, it could be dangerous to predict objective sleepiness at some specific time in a given individual from a single measure of subjective sleepiness.