When we evaluate information, we often believe we are judging the logic of the argument or the accuracy of the facts. However, research across communication, medicine, and data science suggests that the identity and perceived authority of the messenger—the "source credibility"—often play a more decisive role than the message itself.
The Authority Heuristic
In many contexts, the mere act of delivering a message from a trusted source provides a significant baseline benefit that can outweigh the specific content of the communication. Reiter et al. demonstrated this in a landmark randomized controlled trial of a natural language generation system for smoking cessation. They found that the perceived authority of the sender was more influential than the incremental gains from personalizing the message content.
This "authority heuristic" acts as a cognitive shortcut. In complex digital environments, users often lack the time or expertise to verify claims directly. Instead, they use the speaker’s reputation as a proxy for truth. Al Khatib et al. argue that current persuasion research often neglects this factor, showing that a speaker's established authority is just as critical as the argument's content in changing minds. Similarly, Xu found that the prestige of an online reviewer significantly reduces a consumer's perceived risk, as high-status reviewers are viewed as more objective.
The Institutional and Journalistic Premium
While social media has decentralized information, institutional credibility remains a primary driver of how information propagates. An et al. observed that users are far more likely to reshare political news when it originates from established media organizations rather than peer accounts. This "journalistic premium" is further supported by Lee, who found that Americans perceive fact-checking labels from professional news organizations as significantly more effective than those generated by AI or crowdsourcing.
In the realm of public health, this preference for institutions is even more pronounced. Min et al. found that audiences prioritize hospital-verified branding over individual expertise or general media organizations. This suggests that in high-stakes environments, the perceived reliability of a formal institution serves as a critical filter for engagement. Conversely, Li notes that during crises like the COVID-19 pandemic, connecting recognized individual experts with the public is more effective at debunking rumors than issuing anonymous official statements, as it leverages personal credibility.
The Reciprocal Loop of Credibility
Credibility is not a static attribute; it exists in a dynamic, reciprocal relationship with the content being shared. Derczynski and Bontcheva describe this as a feedback loop: a user’s reputation is built on the historical accuracy of their posts, while the perceived truthfulness of a new claim is heavily influenced by that established reputation.
Carchiolo et al. modeled this interaction, showing that user credibility acts as a dual-edged sword—it accelerates the spread of accurate information while simultaneously impeding misinformation. Because users are more likely to transmit news to those who already trust them, credibility becomes the engine of the network's "transmissibility." Hacid et al. emphasize that effective trust-ranking systems must evaluate users and their content simultaneously, as neither can be accurately assessed in isolation.
Behavioral Signaling and the Backfire Effect
Trust is also mediated by "behavioral credibility"—the consistency between a messenger's actions and their claims. Cologna et al. highlight that public trust in climate scientists is heavily influenced by whether the scientists "walk the walk" regarding their own carbon footprints. Discrepancies between a communicator's lifestyle and their professional message can lead to a significant loss of trust.
However, leveraging "liked" or "expert" sources is not a universal solution for persuasion. Spitz et al. discovered a "backfire effect" in a randomized trial: attributing opposing views to liked celebrities or domain experts did not soften polarized opinions. Instead, it often caused participants to harden their original stances. This suggests that when information conflicts deeply with a listener's identity, the source's credibility can be overridden by ideological defense mechanisms.
Vulnerabilities in the Trust Ecosystem
The reliance on source heuristics creates specific vulnerabilities. Leng et al. found that misinformation frequently gains traction by falsely associating claims with reputable figures to bypass critical evaluation. Furthermore, different demographics rely on different trust markers. Lin et al. note that older adults often transfer their trust in traditional, state-regulated media to decentralized social platforms, making them more susceptible to unverified online content.
Ultimately, the effectiveness of any information intervention—whether it is a health app (Flaucher et al.), an agricultural technique (Scientific 2025), or a news aggregator (Laban et al.)—hinges on the perceived integrity of the source. Without a foundation of trust, even the most accurate or diverse information may fail to achieve its intended impact.
Key Foundational Work
Multi-Agent Explanation Strategies in Real-Time Domains (2000) — contributed 1 concept elements to this direction.
Using a Randomised Controlled Clinical Trial to Evaluate an NLG System (2001) — contributed 1 concept elements to this direction.