The Repetition-Based Truth Effect: Substantive and Formal Approaches for Investigating Processing Fluency

The repetition-based truth effect refers to the phenomenon that repeated statements are judged true more often than novel ones. According to the fluency theory, repetition increases the experienced processing fluency, leading to higher truth ratings. In my thesis, I apply substantive and formal approaches to investigate the role of processing fluency in the context of truth judgments. In my first paper, in a more substantive approach, I examined processing fluency by manipulating the syntactic complexity of statements. While I replicated the repetition-based truth effect, syntactic complexity did not affect truth judgments in the first two experiments, which contradicted the predictions derived from the fluency theory. These surprising null effects could only be explained when adding critical auxiliary assumptions, that is, discounting processing fluency as a cue for judging truth. This approach demonstrates the limitations of the fluency theory as a falsifiable scientific theory. In my second paper, I provided a tutorial on multinomial processing tree modeling. This approach is well suited to critically investigate the repetition-based truth effect, as it allows to disentangle contributions of latent cognitive processes. This tutorial demonstrates the benefits of multinomial processing tree models as a formal approach and also facilitates their application to theories of the repetition-based truth effect. In my third paper, I combine substantive and formal approaches by testing the fluency theory using two multinomial processing tree models and receiver operating characteristics. The models implicitly conceptualize the repetition-based truth effect as a response bias towards answering true more often or a reduced discrimination ability between true and false statements. The results support the more counterintuitive conceptualization of the repetition-based truth effect to emerge as a decrease in discrimination ability. Moreover, the results demonstrate that conceptualizing the repetition-based truth effect as a response bias might still be valid given certain boundary conditions. In conclusion, my thesis provides a deeper understanding of the relevance of processing fluency for the repetition-based truth effect. I also highlight the relevance of substantive and formal approaches to systematically examining truth judgments.