Groups spending most of their discussion time on information everyone already knows, while neglecting unique insights held by individuals.
Imagine your family is trying to pick a restaurant. Everyone keeps talking about the pizza place they've all been to, because it's easy to agree on. But your little brother went to an amazing new taco place last week that nobody else has tried. He tries to mention it, but everyone keeps going back to talking about the pizza place because that's what they all know. So you end up at the pizza place again, even though the taco place might have been better.
When groups convene to make decisions, they systematically over-discuss information that all members already possess while under-discussing unique insights held by individual members. This creates a paradox: the very reason groups are assembled—to pool diverse knowledge—is undermined by a collective gravitational pull toward the familiar and already-known. The bias is amplified when groups face time pressure, seek consensus, or work on ambiguous judgment tasks rather than problems with demonstrably correct answers. As a result, groups often arrive at inferior decisions that any single well-informed member could have improved upon, because critical unshared information never surfaces or is dismissed when it does.
The same glitch looks different depending on the terrain. Finance, medicine, a relationship, a team — same mechanism, different costume.
Investment committees tend to anchor discussions on market data and analyst reports available to all members, while overlooking unique sector intelligence or contrarian data points held by individual analysts. This leads to herding behavior and missed opportunities that would have been caught by properly surfacing unshared information.
Diagnostic teams and clinical case conferences disproportionately discuss symptoms and test results available in the shared patient chart, while unique bedside observations or patient history details known to individual clinicians are underweighted or never mentioned, contributing to diagnostic errors.
Garold Stasser and William Titus, 1985, who introduced the hidden profile paradigm in their paper 'Pooling of unshared information in group decision making: Biased information sampling during discussion' published in the Journal of Personality and Social Psychology.
In ancestral environments, information that multiple tribe members independently confirmed was more likely to be accurate and actionable—a form of social verification. Prioritizing commonly held knowledge reduced the risk of acting on one individual's potentially mistaken observation. This consensus-seeking mechanism promoted group cohesion and coordinated action, which were critical for collective survival activities like hunting and defense.
In ensemble AI systems and collaborative filtering, models trained on widely available or popular data points may over-represent common patterns while underweighting rare but informative signals from specialized data sources—a computational analog of shared information bias. In LLM-based group decision support tools, AI summaries may disproportionately highlight information mentioned by multiple participants, reinforcing the bias rather than surfacing unique contributions.
This page taught you the name. The deck turns the name into reflex. 1,100+ swipeable scenarios, 1,100+ defenses, 650+ detection prompts — spaced-repetition Swipe Deck, unlimited Spot-the-Bias Quiz, Defense Playbook, Pre-Flight, My Blindspots, Cheat Sheets, Field Guide e-book. $39.53$59.
