With support from the Digital Democracy Centre (DDC) Thesis Grant, Signe Emilie Hald Stockholm completed her Master’s thesis titled "News Credibility at Risk: The Negative Spill-Over Effects of Misinformation Warnings on Perception Gaps and Credibility Assessments in a High Trust Environment". The Thesis Grant was used to fund data collection for a survey experiment conducted by Epinion.
The thesis investigates the unintended effects of misinformation warnings on the perceived credibility of legitimate news in a high-trust media environment.
You can read more about her interesting project and findings in the abstract below.
Abstract
Warnings about misinformation are widely used to safeguard public discourse, yet emerging research suggests such interventions can carry unintended consequences. This study investigates how exposure to misinformation warnings affects credibility assessments of legitimate news shared on social media in a high trust environment, with particular attention to the mediating role of perceived misinformation prevalence and the moderating effects of alarmist language. Using a Danish survey experiment (N=459), the study randomly assigned participants to one of three groups: a control group, a low-alarm warning condition, and a high-alarm warning condition. Participants then rated the credibility of five authentic news items embedded in a simulated Facebook feed. While no direct effects of the treatments on credibility ratings were found, mediation analysis revealed that warnings significantly increased perceived national misinformation prevalence, which in turn lowered credibility scores across outlets. This indirect pathway was statistically significant, supporting the proposed causal model. Notably, the effect was more pronounced for news media with lower pre-existing public trust, while high-trust outlets such as the public service media DR remained unaffected. Alarm level did not systematically moderate effects, although some media specific effects were uncovered. Media trust emerged as a consistent buffer, attenuating inflated prevalence perceptions and associated credibility losses. The findings suggest that even well-intentioned misinformation warnings can negatively affect credibility assessment of legitimate news through cognitive reframing.