Premium
Combining stock‐and‐flow, agent‐based, and social network methods to model team performance
Author(s) -
Anderson Edward G.,
Lewis Kyle,
Ozer Gorkem Turgut
Publication year - 2018
Publication title -
system dynamics review
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.491
H-Index - 57
eISSN - 1099-1727
pISSN - 0883-7066
DOI - 10.1002/sdr.1613
Subject(s) - computer science , social network analysis , management science , context (archaeology) , diversity (politics) , knowledge management , data science , sociology , engineering , paleontology , biology , world wide web , anthropology , social media
Across disciplines, there has been an increasing interest in combining different simulation methods. Team science provides a particularly challenging context because of the interplay across levels of analysis. For example, team performance is decisively influenced by accumulated individual attributes, the interactions among individuals and emergent team structures—each of which is affected by multiple feedback loops at different levels of analysis. To address these challenges, we compare the modeling methods of stock‐and‐flow models, agent‐based models and social network analysis to argue for the advantages of a hybrid approach to formal mathematical modeling in a team science context. We develop a proof‐of‐concept model, which combines aspects of all three methods, to investigate the effects of expertise, the patterns of members’ interactions and diversity‐based subgroups on team performance. Novel, important insights into team science theory result from this investigation, including, among others, the dynamic tradeoff between diversity and homogeneity on teams’ performance and the importance of the communication network structure in affecting that tradeoff. © 2019 System Dynamics Society