Localizing sensory processing sensitivity and its subdomains within its relevant trait space: a data-driven approach.

Sensitivity arising from enhanced processing of external and internal stimuli or sensory processing sensitivity (SPS) is known to be present in a sizable portion of the population. Yet a clear localization of SPS and its subdomains with respect to other relevant traits is currently lacking. Here, we used a data-driven approach including hierarchical clustering, t-distributed stochastic neighbor embedding (t-SNE) and graph learning to portrait SPS as measured by Highly Sensitive Person Scale (HSPS) in relation to the Big-Five Inventory (neuroticism, extraversion, openness, agreeableness, and conscientiousness) as well as to shyness, alexithymia, autism quotient, anxiety, and depression (11 total traits) using data from more than 800 participants. Analysis revealed SPS subdomains to be divided between two trait clusters with questions related to aesthetic sensitivity (AES) falling within a cluster of mainly positive traits and neighbored by openness while questions addressing ease of excitation (EOE) and low sensory threshold (LST) to be mostly contained within a cluster of negative traits and neighbored by neuroticism. A similar spread across clusters was seen for questions addressing autism consistent with it being a spectrum disorder, in contrast, alexithymia subdomains were closely fit within the negative cluster. Together, our results support the view of SPS as a distinct yet non-unitary trait and provide insights for further refinements of the current SPS concept and scales.