Open a citation
Select a number in Chat to see the cited passage in Sources.
AI Search, Cognitive Bias, and Academic Research
Sources
- Algorithmic Authority Bias in …Paper excerpt · citations 1 and 2
- Processing fluency and perceived credibilitySample source
- Cognitive offloading in research workflowsSample source
Algorithmic Authority Bias in …
… making contexts where users often follow algorithmic recommendations even when those recommendations conflict with their own judgment.
A third key strand is fluency-based trust, also known as a fluency heuristic. In this context, the ease of processing information serves as a powerful indicator of truth and credibility (Oppenheimer, 2008; Reber & Unkelbach, 2010; Hertwig et al., 2008).
In AI-assisted learning environments, this fluency heuristic becomes particularly influential because generative AI responses are typically immediate, grammatically well-structured, and delivered with linguistic confidence. These qualities strongly reinforce perceptions of correctness, even in the absence of verification.
Finally, cognitive offloading refers to the delegation of cognitive tasks to external resources (Risko & Gilbert, 2016).
Chat
Question
Why can a fluent AI answer feel more accurate than it is?
The Fluency-to-Accuracy Illusion
When information is easy to process, people tend to judge it as more credible.
Generative AI responses are often immediate, well formatted, and delivered with confidence.