Can We Assess Trait Authenticity in Others?: Examining the Feasibility of Other-Raters and Automated Video Interviews for Measuring Trait Authenticity

Automated video interviews (AVIs) are increasing in popularity due to their cost-effectiveness, scalability, and convenience, and have been shown to be able to assess personality and other individual difference constructs. At the same time, there has been increasing interest in trait (dispositional) authenticity both in and outside of industrial-organizational psychology, as trait authenticity has been shown to be related to key work and life outcomes, including job performance, work engagement, and well-being. Trait authenticity is currently not assessed in selection despite its relationship with key work outcomes and is currently measured primarily with self-report scales. My research examines the feasibility of using raters and AVIs to assess trait authenticity during selection and whether there are consistent, observable behaviors related to trait authenticity. I found that both raters and AVIs have the potential to assess trait authenticity, with self- and rater-reported trait authenticity showing reasonable convergence (r = .43-.57), and AVI models trained on verbal behavior (RoBERTa embeddings) predicting self- and other-rated trait authenticity at r = .16-.27 and r = .27-.36, respectively. I also found that the use of present focus and 1st person singular were related to high self-reported trait authenticity. Additionally, higher total word count and words per minute were related to high other-rated trait authenticity, while higher average words per sentence and the use of more big words were related to low other-rated trait authenticity. This work contributes theoretically to our understanding of the trait authenticity construct and practically to selection in organizations.