Private Noise and Public Error in Collective Information Acquisition

2026-05-28

AI Summary

This study investigates how communication noise affects the accuracy of collective judgments when groups integrate private evidence and social information to infer an external ground truth, explicitly distinguishing between comprehension noise and production noise for the first time. Through an online experiment involving four-person groups performing 25 rounds of a temperature estimation task, combined with a dynamic updating model and multiple statistical tests, the research reveals that production noise—by generating shared erroneous signals—significantly leads groups to form and sustain incorrect consensus (p = 0.016, 0.025, 0.004), whereas comprehension noise can sometimes facilitate error correction. Notably, GPT-based agents did not exhibit this vulnerability, highlighting a human-specific mechanism in handling uncertainty. The findings demonstrate that noise impairs collective judgment primarily by distorting the reliability of social information rather than amplifying conformity, offering new insights into collective intelligence and information propagation.

Abstract

Collective information acquisition requires groups to combine personal evidence with social information while remaining coupled to the external state. Communication noise can affect this process, but the role of noise remains unclear. In an online experiment, 600 participants worked in four-person human groups estimating a room temperature across 25 rounds while receiving either faithful social information, comprehension noise in which each receiver saw independently perturbed social information, or production noise in which perturbations were stored before display and could be seen by multiple receivers. The thermometer cue was objectively veridical, but its reliability was subjectively uncertain and the unitless 50--250 room-temperature range created a task-induced conflict between displayed evidence and everyday temperature expectations. Production-noise groups spent more rounds tightly clustered around a wrong value than comprehension-noise groups (\(p=0.016\), group-level permutation). Production noise more often created a wrong common signal (\(p=0.025\), Fisher's exact test) and made that signal persist across more rounds (\(p=0.004\), permutation). Dynamic update models showed that production noise was not more harmful because people followed peers more strongly, but because the same peer influence acted on more correlated production-noise perturbations. Exploratory human analyses linked the mechanism to psychological patterns while a GPT-agent experiment clarified a boundary condition: GPT agents registered uncertainty through reduced confidence without reproducing human-scale production-noise vulnerability. Overall, noise did not simply degrade collective information acquisition. Comprehension noise could sometimes improve correction relative to the faithful control, whereas production noise could turn perturbations into common evidence and stabilize consensus on error.

Problem

• collective information acquisition
• communication noise
• production noise
• comprehension noise
• consensus error

Innovation

• production noise
• comprehension noise
• collective information acquisition
• social influence
• common signal error