Peer review is a cornerstone of science, whose quality and efficiency depends on a complex collaboration process, which is sensitive to motivations, incentives and institutional contexts. Recent proofs of failure due to judgement bias, parochialism and misconduct, have contributed to call for reconsideration of the rigor and quality of the process. In this talk, I will present some recent computational studies that looked at behavioural and network effects on the quality of the process. These includes applications of agent-based models that simulate strategic behavior of scientists during the process, and data-driven analysis of a sample of scholarly journals. These works are part of a EU funded project, entitled "PEERE" (www.peere.org), which I lead since 2014. 

‍

http://gecs.unibs.it/squazzoni.html