Abstract
Pupillometry is a popular measure for indexing arousal states. Recent work from the lab revealed that the pupillary effects of cognitive-effort-related arousal multiplicatively interact with luminance: the largest effects occur at low-to-mid luminances (<37 cd/m2), implying a narrow range of conditions ideal for assessing cognitive arousal-driven pupillary differences. In this study, we examine how generalizable this finding is to other forms of arousal. To do so, we assessed how affective arousal-based pupillary modulation interact with luminance. In Experiment 1, participants (n=28) listened to and rated sound clips from the International Affective Digitized Sounds (IADS) of Neutral (low arousal) and Negative (high arousal) valence, while maintaining fixation. To assess pupillary light response functions, the background luminance of the screen cycled through 10 luminance levels (0.92-233.34 cd/m2). In Experiment 2, participants (n=28) viewed and rated images from the International Affective Pictures System (IAPS) of Neutral and Negative valence. The images were presented centrally at fixation (3.8°x2.9°) while the surrounding display cycled through the different luminances. We found that affective arousal interacted with luminance in modulating pupil size differently from cognitive arousal: the effects were smaller and occurred solely at much lower luminances. We also found qualitatively distinct individual differences in modulation by affective arousal, as well as qualitative differences in modulation between the auditory and visual affective stimuli. Critically, there was a sizable group of individuals who did not display any response to the IAPS and IADS stimuli, across all luminances. Taken altogether, our results suggest that arousal is not monolithic: different forms of arousal exert different patterns of effects, suggesting distinct underlying mechanisms. More practically, our findings suggest that high luminances are least ideal when employing pupillometry for arousal, and when selecting ambient luminance levels, one should carefully consider individual differences, as well as the arousal manipulation.