Measuring Arousal: Promises and Pitfalls.

The concept of arousal is ubiquitous and has been suggested as a critical component of many mental and physical phenomena, including emotion, behavior, motivation, learning, attention, motor action, and neural activity. In this review, we discuss theoretical, empirical, and analytic challenges to defining and measuring arousal and suggest avenues for future research to address these challenges. We begin by contesting the typically implicit (but occasionally explicit) assumption in the literature that arousal is a unitary construct representing a generalized, organism-wide phenomenon-an assumption evidenced by the common use of a measure from one arousal domain (e.g., motor action or autonomic activity) as a proxy for arousal in another (e.g., subjective experience). We demonstrate that this assumption is not supported by empirical evidence, and that instead arousal is best conceptualized as multidimensional, comprising multiple factors that can manifest variably both within and across individuals. We then discuss the implications, promises, and pitfalls of adopting this perspective for future research on arousal. We argue that it necessitates employing novel methodological approaches, including highly multimodal data collection from individuals across multiple contexts, preferably in variable real-world settings. It also necessitates bringing innovative data-driven analytic techniques to bear. We conclude that future research stands to dramatically reshape the scientific understanding of arousal by conceptualizing it as multifaceted with a manifestation that is variable in the moment and by leveraging emerging technologies and analyses to improve its measurement.