ABSTRACT
Intellectual humility (IH) has been proposed to be a critical factor facilitating civil discourse. However, wide variability exists concerning what IH actually entails. In this paper, we report a pair of studies investigating how self-ratings and peer-ratings of IH differ in terms of their behavioral associations. In Study 1, students provided round-robin judgments of peers following months of cooperative course work. Self-reported IH was associated with both self-rated and peer-rated openness, whereas peer judgments were primarily associated with self-rated and peer-rated agreeableness. In Study 2, small groups from a community sample were formed for 30-minute conversations about a contentious sociopolitical issue. These recorded interactions revealed that although self-reported IH was associated with high levels of engagement, peer-reported IH was characterized by low negativity and positive, supportive statements. These studies provide evidence that the social and epistemic dimensions of IH are used differently when forming either self or peer judgments.
Acknowledgments
The contents of this paper are solely the responsibility of the authors and do not necessarily reflect the official views of UConn or John Templeton Foundation. We thank Joshua Bias and Linnea Hjelm for their assistance in data collection.
Disclosure statement
No potential conflict of interest was reported by the authors.
Supplementary material
Supplemental data for this article can be accessed here.
Notes
1. Although not empirically validated scales, these sets of items for intellectual humility and intellectual arrogance were included in this study to provide both (1) a comparison point for the GIHS, in terms of how much relative variance is explained by consensus for it relative to this prior indices, and (b) a means of evaluating how consistent rates of consensus of have been across previous studies using round-robin judgments of IH (Meagher et al., Citation2020, Citation2015).
2. The relative target variance was only marginally significant in this study (p <.10). This failure to reach statistical significance is the result of having small group sizes (3–4 people), which reduces the power for this particular analysis (Lashley & Bond, Citation1997; Lashley & Kenny, Citation1998). However, the percentage of relative variance attributable to consensus was greater than 10% and consistent with prior studies (including Study 1), suggesting stability for this estimate.
3. Excluding these covariates from the analyses does little to change the overall pattern of results. The only correlation that ceases to be statistically significant without controlling for these two variables is the relationship between peer-reported GIHS and negative emotion words.