Understanding how moment-to-moment fluctuations in arousal shape cognitive performance is critical for advancing models of attention, decision-making, and mental fatigue. The relationship of arousal and performance has long been of interest in cognitive neuroscience, most famously captured by the Yerkes-Dodson Law, which proposes an inverted U-shaped relationship. The present study examined whether pupil-linked arousal predicts performance, defined by Signal Detection Theory metrics, perceptual sensitivity and response criterion, in an auditory detection task, under low and high cognitive load conditions designed to endogenously modulate arousal. Drawing on the Yerkes-Dodson Law, Adaptive Gain Theory, and Load Theory, we tested whether (H1) perceptual sensitivity follows an inverted U-shaped curve of pupil linked arousal and (H2) response criterion increases linearly with pupil linked arousal within each load condition.

Contrary to both hypotheses, mixed-effects regression models revealed no significant quadratic or linear effects of pupil linked arousal on perceptual sensitivty or response criterion in either condition. While exploratory patterns partially aligned with theoretical expectations, effect sizes were small and unreliable. A likely explanation for the null findings lies in participants’ disengagement from the high load condition which likely undermined the intended arousal manipulation and limited the ability to observe effects.  

These findings highight methodological challenges of reliably manipulating and measuring arousal in dual-task paradigms and underscore the importance of task engagement. Future research should implement real-time feedback for both tasks and enhance load manipulations to better evaluate the nuanced relationship between arousal and cognitive performance