The paradox of partial truths: how text-image congruence shapes social media misinformation dissemination

Abstract

The proliferation of misinformation on social media poses significant societal challenges. While research has burgeoned, the specific role of text-image congruence, the dominant content structure for online content, remains underexplored. This thesis addresses the limited understanding of how text-image misinformation often operates through “partial truths,” where the interplay and congruence between text and image, rather than outright falsehoods in either modality, shape persuasion and dissemination dynamics. It investigates how text-image congruence influences these processes through conceptualization, effects analysis, and dynamic explorations.
Employing a multi-method approach across three interconnected studies, this research first develops a comprehensive conceptualization of text-image congruence (Study 1). Through systematic literature review and grounded theory analysis, four key dimensions (i.e., Object, Semantic, Emotion, and Style congruence) are identified, grounded in the information process perspective. An integrative framework based on the S-O-R paradigm is proposed, synthesizing multiple persuasion theories to elucidate the cognitive and affective mechanisms driving multimodal persuasion.
Second, the thesis empirically investigates the direct effect of text-image congruence on real-world misinformation dissemination (Study 2). Based on the theorization from a narrative experience perspective, analysis of observational data reveals significant inverted U-shaped relationships between both object congruence and semantic congruence with misinformation dissemination, indicating the moderate congruence levels optimally balance narrative immersive and participating experiences to facilitate spread.
Third, adopting a configurational perspective, the thesis explores the complex interplay of influencing factors driving high dissemination (Study 3). A sequential mixed-methods design, combining qualitative interviews and fuzzy-set Qualitative Comparative Analysis (fsQCA), identifies multiple antecedents and reveals four distinct configurational pathways (i.e., Visual-abundance-centric, Object-recognition-centric, Authority-centric, and Simplicity-centric) leading to high text-image misinformation dissemination.
Collectively, this thesis illuminates the “paradox of partial truths” by advancing multimodal communication theory within IS by providing a nuanced, multi-dimensional understanding of text-image congruence, demonstrating non-linear congruence effects, and revealing the complex configurational dynamics underlying misinformation spread. The findings offer significant theoretical contributions and practical implications for platform design, media literacy, and regulatory policy aimed at mitigating the impact of compelling but misleading multimodal narratives in digital ecosystems.

Date of Award	15 Nov 2025
Original language	English
Awarding Institution	University of Nottingham
Supervisor	Boying Li (Supervisor), David Phang (Supervisor), Boying Li (Supervisor) & David Phang (Supervisor)