Beyond the text: Voice as a stable marker of self-esteem

Abstract

Self-esteem is a core personality trait, yet its assessment often relies on self-reports vulnerable to contexts. Integrating Trait Activation Theory, Whole Trait Theory, and the Self-Organizing Self-Esteem model, this study explores voice as a context-independent marker of self-esteem. Across three tasks (N = 211) varying in self-relevance and social evaluative threat, voice features outperformed text in recognizing self-esteem and generalized across situations. Specifically, greater loudness, higher and more stable warmth, and shorter and less variable silence durations indicated higher self-esteem. Based on these findings, we propose the Modality-Based Whole Trait Theory, extending existing theories by emphasizing modality as a boundary condition for personality expression. These findings advance objective personality assessment and highlight the potential of voice for capturing trait-level information.