Compassion is partially selfless: Public self-concept accessibility following compassion and threat

ABSTRACT

Drawing from the largely separate literatures on compassion and the self-concept, we hypothesized that compassion would lead to literal selflessness in the form of reduced cognitive accessibility of the public self-concept. Across two studies, we found preliminary support for this hypothesis. Study 1 (N = 154) found that a compassion induction, relative to control conditions, reduced accessibility of the public self, but did not impact accessibility of the true self. Study 2 (N = 172) employed methodological refinements and found that compassion reduced accessibility of the public self, but only when preceded by a threat manipulation. These studies’ implications and possible explanations for their apparent inconsistencies in findings are discussed. Taken together, these studies provide partial support for a unique causal link between compassion and reduced cognitive accessibility of the public self.

KEYWORDS:

• Compassion
• self
• public self
• true self
• prosocial
• self-awareness

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Though prior research has commonly referred to the public self as the “actual self,” Schlegel et al. (2009) noted that the term public self may be more appropriate. For this reason, and also because we believe it affords greater clarity in the context of the present research, we use the term public self.

2. Alternative means of classifying these overlapping attributes gave largely similar results: the two-way interaction between compassion and self-concept remained significant when overlapping words were completely excluded from the model, B = 1.13, p = .01, and remained marginally significant when overlapping words were included in the public self instead of the true self, B = −.40, p = .055.

3. There were no significant differences among the three control conditions: imagine-self perspective taking, self-compassion, and distraction (all ps >.72).

4. We conducted analyses to evaluate a possible alternative interpretation of the two-way interaction between self-concept and IOPT – namely, that this pattern of findings could be confounded by the valence of the attributes comprising participants’ component self-concepts. The interaction between compassion and self-concept remained significant (B = 1.13, p = .02, 95% CI [0.16, 2.11]) even when controlling for the interaction between compassion and valence.

5. This dot probe task consisted of 24 practice trials and 96 test trials. Test trials featured neutral and angry face images (Langner et al., 2010) presented in random order.

6. Similar to Study 1, the two-way interaction between self-concept and IOPT within the threat condition remained significant when controlling for valence (b = 0.38, p = .01, 95% CI [0.09, 0.67]).

7. There was no significant difference between the two control conditions: NPT and distraction (p = .17).