In infancy, it’s the extremes of arousal that are ‘sticky’: Naturalistic data challenge purely homeostatic approaches to studying self-regulation