Toward a Structurally-Sound Model of Uncertainty-Related Personality Traits

Abstract

Numerous personality constructs have been proposed over the last 70 years to describe individual differences in attitudes toward uncertainty, including Intolerance of Ambiguity (IA), Intolerance of Uncertainty (IU), Uncertainty Orientation (UO), Need for Closure (NFC), and Personal Need for Structure (PNS). Despite differences in their theoretical foundations and applications, these constructs share considerable similarities in conceptualization and measurement, which have confounded researchers about how they can be meaningfully distinguished. Study 1 examined the facet- and construct-level overlap among multiscale measures of IA, IU, and NFC—constructs that have received the lion’s share of research attention—also including the Uncertainty Response Scale (URS; Greco & Roger, 2001). Confirmatory factor analysis revealed that the general factors underlying each measure were identical or near-identical with one another (in two US samples but not an Italian sample), and exploratory factor analyses of both the items and facets of each measure found that only four factors replicated across US and Italian samples. Moreover, the four factors uncovered showed strong concurrent validity with trait measures of neuroticism, orderliness, curiosity/exploration, and naïve epistemic beliefs, which suggested that they reflect distinct types of responses to uncertainty, consistent with the model assumptions of the URS. Another two studies examined the discriminant predictive validity of the four factors with respect to evidence gathering (Study 2) and avoidance of ambiguity under conditions of threat and high cognitive demand (Study 3). Their results confirmed that these factors predict different behavioral outcomes and have more predictive power than NFC, IA, and IU scales. Overall, this work demonstrates that a four-factor model has psychometric and conceptual advantages over current measures of uncertainty-related constructs, which are lacking in discriminant validity.