Sanjeev Goyal (Cambridge)
"Learning in Canonical Networks (with Syngjoo Choi, Fred Moisan and Tony To) "
Abstract: We conduct experiments on the binary signal binary action observational learning problem on networks. Agents initially guess the binary state of the world after observing a private signal. In subsequent rounds, agents observe their neighbors' previous guesses before guessing again. How does network structure shape information aggregation? We consider groups of 10 and 40 located in social networks — Erdos-Renyi, Stochastic block (reflecting homophily), royal family (reflecting `influencers') and RGG (containing clans). We show that networks have powerful effects on social learning -- whether the groups converge to consensus and if they do whether they choose the correct or the wrong guess. We study decision rules that can account for individual behaviour in these networks.
When: Monday 14th June 2021 - 1:00pm
Where: Zoom
Reading Group: Theory Workshop