You’re Not Wrong, You’re Just Loud: Dunning-Kruger in Group Chats
Dr. Daphne Voxman, Department of Digital Overconfidence Studies, Western School of Misapplied Psychology (WSMP)
Abstract
In the age of digital discourse, group chats have become both the agora and the battleground of modern social interaction. Yet beneath the emojis, GIFs, and unsolicited political takes lies a quiet epidemic of epistemic audacity. This paper investigates the manifestation of the Dunning-Kruger effect in group chats: the phenomenon whereby individuals with the least knowledge often speak with the most confidence—and the highest frequency. Drawing on transcript analysis, volume-to-accuracy ratios, and several ruined friendships, we construct a profile of the Chat Confidant: a digital figure who knows little, says much, and always replies first. Our findings reveal a direct correlation between low subject expertise and high text volume per topic, especially in the domains of economics, climate change, and why restaurants are “actually a scam.”
1. Introduction: Welcome to the Thread
The modern group chat is a microcosm of unmoderated democracy: flat, fast, and fertile for confusion. It is where plans are made, memes are shared, and knowledge is distorted beyond recognition. Anyone who has spent more than ten minutes in a group chat has encountered the confident wrongness of the Chat Confidant—a person who, though utterly unqualified, nevertheless takes the lead in every debate.
This is not a new phenomenon, but its intensity has surged in the digital age. What once required a barstool or a captive dinner table now only demands thumbs and Wi-Fi. The group chat, in short, has become the ideal greenhouse for Dunning-Kruger to bloom.
2. Methodology
We analyzed 112 active group chats across platforms including WhatsApp, Facebook Messenger, Discord, and encrypted doomsday forums. Chats ranged from family circles to college alumni groups, and one particularly chaotic book club thread that had not discussed a book since 2019.
Each message was coded based on its content category (e.g., factual assertion, unsolicited opinion, emotionally charged rant, passive-aggressive correction) and assessed for confidence-to-corroboration ratio (CCR). We then mapped individual participant expertise levels (self-reported and peer-reviewed) against contribution volume and tone.
We also surveyed 300 chat participants on their perception of who “knows the most” in their group and cross-referenced these results with actual accuracy rates based on third-party fact-checking services and Wikipedia at 2 a.m.
3. Findings: Loudness as Epistemic Camouflage
3.1 The Confidence-Inversion Curve
Our data reveal an unmistakable pattern: those with the least actual expertise in a topic tend to speak with the most unwavering confidence. This confidence often manifests within seconds of a topic being raised. In 68% of political discussions, the first responder was later rated as the least informed by external evaluators—but their tone was unanimously described as “borderline presidential.”
This phenomenon, which we term the Confidence-Inversion Curve, is a hallmark of Dunning-Kruger group chat dynamics. As confidence goes up, veracity plummets.
To visualize this, we present a scatter plot comparing self-reported confidence and independently assessed accuracy across participants in several active group chats. As shown in Figure 1, individuals such as “Todd” and “Ben” exhibit very high confidence scores while scoring significantly lower in factual accuracy. Conversely, participants like “Zoe” and “Maya” display modest confidence but high accuracy. The inverse correlation is both clear and alarming.
Figure 1. Scatter plot mapping confidence vs. accuracy among eight typical group chat participants. Note the steep inverse relationship between confidence and correctness.
3.2 Subject Matter Vulnerability
Certain topics are particularly susceptible to confident wrongness. Chief among them are macroeconomics (especially inflation, despite no one having studied it), epidemiology (particularly among people who “watched a documentary once”), nutrition (including but not limited to detox tea, bone broth, and “processed vibes”), law (especially when someone in the chat has been to small claims court), and parenting (whether or not the participant has children).
In one striking instance, a participant in a group titled “Ski Trip 2025” provided a 2,300-word monologue on how vaccines alter DNA, prompting the group’s ski instructor to delete the app.
3.3 The “Actually” Threshold
The word “actually” was found to be a reliable predictor of Dunning-Kruger manifestation. Phrases such as “Actually, if you think about it…” or “Actually, that’s a myth” were used 4.6x more often by low-expertise contributors. These statements were often paired with links to articles that had either been debunked, were hosted on sites with .biz domains, or were screenshots of screenshots.
4. Behavioral Taxonomy: Species of Chat Confidants
Through cross-chat ethnography, we identified four primary archetypes:
The First Responder – Types before they think. Always responds within ten seconds. Uses phrases like “Simple: just…” or “You’re overcomplicating it.”
The Wiki-Warlord – Screenshots a Wikipedia paragraph and treats it as divine law. Often pastes it without reading the last sentence, which contradicts their point.
The Link Baiter – Sends articles without commentary. When pressed, responds “I didn’t read it yet, but the headline is wild.”
The Theorist – Has an original explanation for everything. Economics? “We’re just using money wrong.” History? “What if the Romans never existed?”
Each type exhibits elevated CCR scores and thrives in low-moderation environments.
5. Psychological Drivers
Why do so many people speak so confidently about things they do not understand? Our interviews suggest a cocktail of factors:
Status Maintenance: In digital tribes, assertion is often mistaken for knowledge. The loudest voice earns the perceived role of “thought leader,” regardless of substance.
Information Paranoia: Many individuals now feel that traditional expertise is part of a larger deception. They believe their “gut instincts” to be more pure than peer-reviewed evidence.
Immediacy Culture: Group chats reward speed over accuracy. The first voice heard often shapes the tone, creating a false sense of dominance.
Confirmation Comfort: Dunning-Kruger thrives on echo. Group chats often lack dissent, turning them into soft bunkers for self-assured misinformation.
6. The Cost of Loudness
Beyond minor embarrassment and digital eye rolls, there are deeper consequences. Plans derail, debates fester, and real information gets buried under layers of confident wrongness. In one alarming case, a friend group missed a flight because the Chat Confidant insisted “you don’t need passports for Canada.” They were American.
In family chats, medical advice from non-medical relatives led to three misdiagnoses, two garlic-based treatments, and one lawsuit. In professional contexts, chat-based overconfidence cost one startup $18,000 after a marketing manager insisted they “didn’t need a lawyer to trademark stuff.”
Misinformation spreads like wildfire, and the person holding the match often has the fewest matches left in their trivia bank.
7. Recommendations for Group Chat Hygiene
Introduce a Slow Mode feature for heated topics. Let replies require a two-minute cool-off period to encourage reflection.
Assign a rotating Devil’s Fact-Checker, whose only job is to gently deflate nonsense.
Add a Confidence Slider to each message: users must rate their certainty from 1 (wild guess) to 10 (I wrote the textbook).
Create a group chat manifesto that includes the clause: “We are all capable of being wrong. Especially Todd.”
8. Conclusion: Typing Isn’t Thinking
The group chat, for all its community potential, has become a playground of poorly sourced certainty. The Dunning-Kruger effect thrives in these spaces, elevating the loudest and ignoring the nuanced. But by recognizing this dynamic, we can reclaim our digital habitats—not by silencing, but by slowing down.
To know less is human. To shout regardless is modern.
Let us instead aim for a humble emoji, a clarifying link, and perhaps even the rarest group chat gesture of all: the admission, “I don’t know.”
References
- Voxman, D. (2023). Typing Isn’t Thinking: A Cognitive Cartography of Group Chats. WSMP Press.
- Dunning, D. & Kruger, J. (1999). Unskilled and Unaware of It: How Difficult It Is to Detect One’s Incompetence.
- ThreadWatch Analytics (2022). Confidence-to-Corroboration Metrics in Digital Messaging.
- ChatForensics.biz (2021). Actually: The Most Dangerous Word in Casual Discourse.
- GroupText Quarterly (2023). From LOL to Litigation: The Hidden Costs of Group Chat Advice.
- Blurt, E. (2023). I Said It First, Therefore It’s Right. A Memoir.
- Thistlewhisker, B. (2022). Digital Narcissism in Micro-Communities.
- Wikipedia (all of it, apparently).
Disclaimer: This article is a work of satire. Please don’t cite it in your next group chat argument—unless it’s with Todd.