A general factor for trust?: Testing latent factor structures of trust across institutional and interpersonal contexts.

Abstract

The literature is replete with multi-dimensional self-report assessments of trust. It is not clear whether these dimensions are statistically distinguishable across institutional and interpersonal contexts, respectively. We sought to provide empirical insights that might permit researchers to refine the conceptualisation and dimensionality of trust, as well as provide suggestions for institutions or individuals hoping to cultivate trust. Specifically, we aimed to test whether evidence for a general trust factor would emerge in relation to trust in institutions and other people. 588 adults completed an online survey assessing dimensions of trust measured in institutional and interpersonal contexts. Confirmatory factor analyses (CFAs) revealed that a 'many-factor' correlated model provided the best fit statistics in both interpersonal and institutional contexts. Higher-order and bi-factor models also produced excellent fit. Exploratory structural equation modelling (ESEM) revealed a high-degree of item cross-loadings, suggesting that the tested trust dimensions were not as distinct as predicted. The bi-factor ESEM model found that all items had significant loadings on a general factor, supporting the notion of a general trust factor. This effect appeared more persuasive in interpersonal contexts, relative to institutional contexts. Trust-related dimensions may not merely be distinct, correlated constructs. Statistical evidence produced in the current study aligns with the suggestion that people's trust-related perceptions may, in part, be influenced by a general factor. We propose the theory of epistemic trust as a candidate for interpreting the general trust factor.