Cognitive biases are systematic patterns in how humans process information, often leading us to deviate from rational judgment. In the misinformation context, cognitive biases are foundational mechanisms that make people vulnerable to false claims and resistant to correction.
Key biases relevant to misinformation
Confirmation bias
The tendency to seek, interpret, and recall information in ways that confirm what we already believe. If you want to believe something, you go looking for evidence for it — and it feels like thinking, but it's actually just filtering. Confirmation bias is amplified by:
- Partisan media (cable TV, social media algorithms) designed to tell you what you want to hear
- Echo chambers and filter bubbles, where users are surrounded by like-minded people
- Availability heuristic (we overweight examples we easily recall, and algorithms feed us examples repeatedly)
Backfire effect
When presented with evidence contradicting a belief, people sometimes strengthen their false belief instead of updating. This has profound implications for fact-checking: a badly designed correction can be counterproductive. However, empirical research suggests the backfire effect is smaller than initially feared and can be mitigated by:
- Presenting corrections in a friendly, non-hostile tone
- Providing alternative explanations (not just saying what's false)
- Repeated exposure over time
Dunning-Kruger effect
People with low knowledge often overestimate their competence. Relatedly, people who hold false beliefs are often harder to persuade than people who are simply uninformed — the misinformed think they already know, while the uninformed know they need to learn. This reverses the intuition that correction is easier than education.
Motivated reasoning
People evaluate evidence differently depending on whether accepting it supports or threatens their identity, values, or group membership. Claims that align with political affiliation, religious belief, or group identity are accepted with lower evidence thresholds; claims that threaten group identity are subject to much higher scrutiny. This is not irrationality but identity-protective cognition.
Illusory truth effect
Repeated exposure to a claim makes it feel more true, even when people know the source is unreliable. Advertisers and propagandists exploit this: if a false claim is repeated enough, it becomes familiar, and familiarity breeds (false) truth.
Psychological vs. structural explanations
Cognitive biases are wired in — they are features of the human mind that evolved for good reasons (e.g., confirmation bias helps us avoid constantly second-guessing ourselves). The key insight is that biases are not created by modern media but are amplified by information structures that exploit them:
- Without the internet: Confirmation bias existed, but you had to actively seek out like-minded media
- With the internet and algorithms: Confirmation bias is auto-fed — the algorithm learns what keeps you engaged and serves you more
Addressing bias
Individual level: Awareness, intellectual humility, active reasoning against your inclinations.
Social/institutional level:
- Diverse information diets and exposure to viewpoint diversity (though this can backfire if framed as "balance" between truth and falsehood)
- Respectful, non-threatening presentation of corrections
- Fact-checking and source credibility cues, though effects are modest
- Prebunking and inoculation (pre-emptively explaining persuasion techniques before people encounter them)
Systemic level:
- Media design that promotes epistemic diversity without false balance
- Algorithmic transparency and design constraints that don't optimize purely for engagement
- Institutional credibility in science, media, and government
Key videos
- Post-Truth in the Internet Age — McIntyre discusses confirmation bias, backfire effect, and Dunning-Kruger effect as foundations of post-truth
Related topics
- Misinformation — false claims exploit cognitive biases
- Echo chambers and filter bubbles — information structures that amplify biases
- Intervention strategies — inoculation and prebunking approaches to address biases
- Belief in misinformation — conditions under which people form and maintain false beliefs
Open questions
- How large is the true backfire effect, and under what conditions does it occur?
- Can people be taught to recognize their own confirmation bias, or is the bias itself inoculated against critique?
- Do biases operate differently across domains (politics, health, science)?
- What is the minimal set of cognitive debiasing interventions that scale to populations?
