Investigating the stability of conspiracy mentality versus specific conspiracy belief : a longitudinal perspective : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Psychology at Te Kunenga Ki Pūrehuroa - Massey University, Albany, New Zealand

Abstract

Conspiracy mentality and specific conspiracy beliefs are often treated interchangeably, yet theories suggest that they differ in both conceptual scope and temporal stability. Conspiracy mentality is thought to reflect a broad, trait-like worldview, whereas specific conspiracy beliefs arise from this mentality and are more closely tied to particular events or actors. Imhoff and colleagues argue that a conspiracy mentality should be more temporally stable, given its similarity to a personality trait, while specific conspiracy beliefs should fluctuate with changing contexts. The present study directly compared the stability of these constructs using 25 monthly waves of longitudinal survey data collected from 986 participants across Australia, New Zealand, and the United Kingdom. Conspiracy mentality was measured using the Conspiracy Mentality Questionnaire, and specific conspiracy beliefs were assessed using 11 contemporary items. Stability was evaluated using descriptive analyses, intraclass correlations (ICCs), which quantify the proportion of total variance attributable to between- versus within-person differences, and a multilevel structural equation model (MSEM) that separates within-person fluctuations from stable between-person differences. Both constructs showed high temporal stability at the item and latent levels. Item-level ICCS ranged from .73 to .83 for conspiracy mentality and from .76 to .86 for specific conspiracy beliefs, with average ICCs of .88 and .93, respectively. MSEM variance estimates similarly indicated that most variance occurred between individuals rather than within individuals over time. Conspiracy mentality showed a latent ICC of .91. In contrast, specific conspiracy beliefs demonstrated even greater stability at .96. These patterns held when CMQ responses were rescaled to match the specific conspiracy belief response format. Across analyses, specific conspiracy belief exhibited slightly higher stability than conspiracy mentality, suggesting that an individual’s endorsement of specific narratives is at least as trait-like as the broader conspiratorial worldview. These findings challenge the assumption that specific conspiracy beliefs are more context-dependent and malleable than general conspiracy mentality. Instead, both constructs appear highly stable over time. Limitations include the low endorsement of 3 some specific beliefs, the two-year timeframe, and the possibility that the findings reflect features of the measures rather than the constructs. Nevertheless, there are practical implications for measurement, theory, and intervention design.