Ce document est lié à :
info:eu-repo/semantics/altIdentifier/arxiv/2003.06160
Ce document est lié à :
info:eu-repo/semantics/altIdentifier/doi/10.31234/osf.io/bufjk
Ce document est lié à :
info:eu-repo/semantics/altIdentifier/pmid/32900307
http://creativecommons.org/licenses/by/ , info:eu-repo/semantics/OpenAccess
Bertrand Jayles et al., « The impact of incorrect social information on collective wisdom in human groups », HAL-SHS : économie et finance, ID : 10.31234/osf.io/bufjk
A major problem resulting from the massive use of social media is the potentialspread of incorrect information. Yet, very few studies have investigatedthe impact of incorrect information on individual and collective decisions.We performed experiments in which participants had to estimate a seriesof quantities, before and after receiving social information. Unbeknownstto them, we controlled the degree of inaccuracy of the social informationthrough ‘virtual influencers’, who provided some incorrect information.We find that a large proportion of individuals only partially follow thesocial information, thus resisting incorrect information. Moreover, incorrectinformation can help improve group performance more than correct information,when going against a human underestimation bias. We thendesign a computational model whose predictions are in good agreementwith the empirical data, and sheds light on the mechanisms underlyingour results. Besides these main findings, we demonstrate that the dispersionof estimates varies a lot between quantities, and must thus be consideredwhen normalizing and aggregating estimates of quantities that are verydifferent in nature. Overall, our results suggest that incorrect informationdoes not necessarily impair the collective wisdom of groups, and can evenbe used to dampen the negative effects of known cognitive biases.