Cognitive, endocrine and mechanistic perspectives on non-linear relationships between arousal and brain function.

It has been almost a century since the first paper describing a non-linear relationship between arousal and behavioral performance was published (Yerkes and Dodson 1908). This study, an analysis of the influence of task difficulty and stress on discrimination learning in the dancing mouse, stands apart from all others. The paper was published without statistical analyses (statistics had not yet been conceived) and with sample sizes as small as 2 mice per group (an unacceptably low level of power by modern standards). Despite the limitations of their study, the findings of Yerkes and Dodson were subsequently replicated in cats (Dodson, 1915), rats (Broadhurst, 1957; Telegdy and Cohen 1971) and people (Dickman, 2002; Bregman and McAllister 1982; Anderson, 1994), and became part of the lexicon of the field of psychology as the “Yerkes-Dodson Law” (Young, 1936; Eysenk, 1955). In brief, Yerkes and Dodson found that when mice were given a simple discrimination task their performance improved linearly with increases in arousal. With more difficult tasks, the performance of the mice improved with moderate with increases in arousal, but at the highest levels of arousal their performance was impaired, forming an overall non-linear (inverted-U) shaped relationship between arousal and performance. This task-dependent influence on the shape of arousal-performance curves is illustrated in Figure 1.