Clarifying the Relationship Between Ostracism and Relational Devaluation

Abstract

We examine how three perspectives on relational devaluation relate to needs threat following ostracism. In two experiments with 179 first-year psychology students, distress was greatest when participants were ostracized without any prior throws, and distress decreased linearly with increasing prior inclusion. In Experiment 3, using 76 first-year psychology students, we manipulated expectations of exclusion and found expectations predicted distress following ostracism, suggesting ostracism's distress can be influenced by norm-based expectations of inclusion, and that progressive relational devaluation is not a necessary condition for ostracism's distress.

Keywords:

• expectations
• ostracism
• rejection
• relational devaluation

Notes

1. The closest exception is the constant rejection condition in Buckley et al. (2004), but in this condition the person received ambiguous feedback, that the listener wanted to get to know them only 2 on a scale of 1 to 9, which could still be interpreted as the listener slightly wanting to know the participant. Using the endpoint of the scale would have been clearer.

2. Except for face validity, the needs threat scale has never been demonstrated to have four distinct subscales. A maximum-likelihood factor analysis of this scale for all the participants in this current study (N = 255) suggested a one-factor solution was most appropriate, with the first factor accounting for 53% of the variance, and the next factor's eigenvalue being close to unity, 1.2. Separate MANOVAs conducted on each study confirmed this interpretation. All subscales had similar effects in Study Two, Box's M = 64.77, p = .24, and Study Three, Box's M = 41.25, p = .18. The subscales showed minor variations in Study One, Box's M = 67.13, p = .001, with those being ostracized after ten throws feeling slightly less meaningful existence and slightly higher self-esteem than those ostracized after two throws. This result was not of great theoretical interest and could very likely be a Type I error.

3. The expectation manipulation influenced both personal expectations of exclusion and also normative expectations of inclusion. As such, expectations in our context include both personal and normative components.