People display systematic priorities to self-related stimuli. As the self is not a unified entity, however, it remains unclear which aspects of the self are crucial to producing this stimulus prioritization. To explore this issue, we manipulated the valence of the self-concept (good me vs. bad me) — a core identity-based facet of the self — using a standard shape-label association task in which participants initially learned the associations (e.g., circle/good-self, triangle/good-other, diamond/bad-self, square/bad-other), after which they completed shape-label matching and shape-categorization tasks, such that attention was directed to different aspects of the stimuli (i.e., self-relevance and valence). The results revealed that responses were more efficient to the good-self shape (vs. other shapes), regardless of the task that was undertaken. A hierarchical drift diffusion model (HDDM) analysis indicated that this good-self prioritization effect was underpinned by differences in the rate of information uptake. These findings demonstrate that activation of the good-self representation exclusively facilitates perceptual decision-making, thereby furthering understanding of the self-prioritization effect.

The neural representation of self is a fundamental question for brain research. Employing activation likelihood estimation (ALE) meta-analyses, we assessed the commonalities and distinctions between different components of the self by focusing on the ‘physical’ self and the ‘psychological’ self – assessed respectively through face processing and self-referential tasks. We first conducted ALE meta-analyses by computing the convergence of findings on brain activation in self-face recognition and self-referential studies respectively. Contrast and conjunction analyses of these two meta-analytic results were then applied to extract the distinctions and commonalities in self-face and self-reference tasks. Facial self processing was particularly associated with lateral brain regions with a right hemispheric dominance, while processing psychological self predominantly activated cortical midline structures, more specifically the anterior cingulate cortex/superior frontal cortex. In contrast, the conjunction analyses showed that the two aspects of self-processing recruit the dorsal anterior cingulate cortex and the left inferior frontal gyrus extending to the insula. A framework including both distinct and common neural representation of selfs is discussed.