Dimensionality, interconnectedness, and cross-national comparability : studies of the global trust in multinational representative samples : a thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy in Psychology at Massey University, Albany campus, Aotearoa New Zealand.

Abstract

Trust has been commonly portrayed as a desirable characteristic for both individuals and societies. However, debates around the conceptualisation of trust are still ongoing, as recent literature has challenged the conventional treatment of social and political trust as two unidimensional and separate constructs. It is believed that a simplistic conceptualisation and measurement of trust may overlook the multifaceted and interconnected nature of trust and potentially distort cross-national comparisons. This thesis investigates the dimensionality, interconnectedness, and cross-national comparability of trust using representative samples from a multinational online survey project. Study 1 demonstrated a conditional interconnection between social and political trust in the context of 11 democratic societies: different types of trust formed two clusters that centre around social and political trust, respectively, but they were interconnected through the specific links between trust in neutral (non-partisan) institutions on the one hand and trust in community on the other hand. Study 2 demonstrated a culturally and politically contingent view of the structure of trust through a confirmatory factor analysis of the Global Trust Inventory. In four East Asian societies, two different models of trust (China model and Democratic East Asian model) fit the data better than the model suitable for the 11 western democracies in Study 1 (Western model), probably due to differences in culture and political systems. Study 3 demonstrated that two sub-measures of the Global Trust Inventory, capturing two types of social trust, were metrically invariant across 18 culturally and politically heterogeneous societies and across a six-month time interval. Results of a cross-lagged panel analysis further suggested that there was a bidirectional link between trust in community and life satisfaction, but life satisfaction was longitudinally associated with trust in close relations, not vice versa. Overall, this thesis supports a multidimensional and conditionally interconnected view of trust and explores ways of dealing with measurement non-invariance in cross-national survey research.