The proliferation of misinformation has profound impacts on public discourse and decision-making. This dissertation explores the effectiveness of misinformation corrections, focusing on how message, source, and audience features of corrections influence responses. Specifically, with perceived language intensity of the correction as a central message feature, the study examines its unique and interactive effects with other factors in influencing correction effectiveness, including source credibility as a source feature, and attitude discrepancy and issue involvement as audience features. Drawing on the language expectancy theory (LET) and the information processing model, the Language Intensity and Correction Effectiveness Model (LICE) is proposed and examined under different contexts—for familiar and unfamiliar sources of corrections—where source credibility plays distinct roles in the processes.
To test the model, an experimental design was implemented in which participants were exposed to pro-attitudinal misinformation followed by corrections with varying levels of language intensity. These corrections were delivered by sources of differing familiarity and credibility. Audience features such as prior beliefs in the veracity of the misinformation and issue involvement were measured. Correction effectiveness was operationalized as reduction in belief in the misinformation, by comparing perceived credibility of the misinformation after exposure to the correction to the control group with no correction, and between experimental conditions with varying levels of influencing factors. Message and source credibility perceptions of the correction, perceived language intensity of the correction, and demographics were also measured.
Findings suggest that corrections are generally effective in reducing belief in the misinformation regardless of language intensity, but high-intensity language has the potential to undermine correction effectiveness across varying levels of source credibility, attitude discrepancy, and issue involvement. Specifically, intense language negatively impacts correction effectiveness through decreased message credibility of the correction, which in turn decreases source credibility for unfamiliar sources. The detrimental effect of high-intensity language remains even for familiar sources with high prior credibility perceptions, and is particularly pronounced among individuals with low attitude discrepancy with the correction and high involvement with the issue.
Overall, this research reveals a distinct process in responses to misinformation corrections, where individuals critically examine all possible cues to assess credibility, emphasizing a balanced approach to consider the combined effects of message, source, and audience features to develop more targeted and effective correction messages for different source and audience groups.