Caffeine and Cognitive Performance: The Nonlinear Yerkes–Dodson Law

This study presents a test of the Yerkes–Dodson Law (YDL; Yerkes and Dodson, 1908), which is understood to predict a negative quadratic relationship between arousal and performance (‘inverted‐U’ hypothesis), and a lower level of arousal for optimal performance on more difficult tasks than easier tasks (‘task difficulty’ hypothesis). A number of recent studies (e.g. Neiss, 1988) have questioned the validity of the YDL on several grounds: the confusion of theory and model; observed linear arousal–performance relationships; non‐specific definitions of arousal; and poor experimental design. A single‐blind modified version of Anderson's (1994) within‐subjects study (N = 10) was performed, utilizing graded cortical arousal manipulations of caffeine (100 mg cumulative dosages to a maximum of 600 mg), and four tests of basic cognitive ability in the procedural alphanumerical domain (with counterbalancing of drug/placebo session and ordering of presentations of tasks). The ‘inverted‐U’ hypothesis was supported in three out of four experimental conditions (easy and difficult numerical, and difficult alphabetical tasks; p < 0·05). No support was found for the task difficulty hypothesis. The results are discussed in terms of the emergence of nonlinearity in neural–cognitive interactions as a fundamental quality of drug–behaviour interactions. © 1997 John Wiley & Sons, Ltd.