THEMIS.COG: Theoretical and Empirical Modeling of Identity and Sentiments in Collaborative Groups

Dartmouth College, Hanover NH

Investigators

Abstract

The proposed research will provide new insights into the dynamics of self-organized collaborations, in which people come together organically to work on a common problem, without prompting by a third party. Understanding the social forces behind self-organized collaboration is increasingly important in today's society. Technological and social innovations are increasingly generated through informal, distributed collaboration processes, rather than in formal, hierarchical organizations. This project will use a data-driven approach to explore the social and psychological mechanisms that motivate self-organized collaborations and determine their likelihood of success or failure, focusing on the example of open, collaborative software development in online collaborative networks like GitHub (github.com). We offer a mathematically precise model for predicting and testing collaborative dynamics, which builds on a long tradition of sociological theory and research. By accounting for noise and uncertainty in processing social events, this work will make it possible to study the reach of social psychological mechanisms to forms of interaction other than face-to-face communication, where interpretations of events are far less certain, and to cross-cultural interactions. The research builds on a powerful sociological theory of group processes known as affect control theory (ACT) and a recent probabilistic generalization of it, Bayesian affect control theory (BayesACT). BayesACT applies insights from Bayesian probability theory to explain how people learn and adjust meanings through social experience, and show how stable interaction dynamics emerge from individuals' uncertain and noisy perceptions of their own and others' identities. The model rests on the idea that humans are motivated in their social interactions by affective alignment: They strive for their social experiences to be coherent at a deep, emotional level with their sense of identity and the cultural beliefs they share with others. BayesACT makes explicit predictions about online interactions in a collaborative group, based on the notion that each group member holds an identity that is learnable, mathematically describable, and complementary to those of other group members. These predictions allow for deeper and more focused data mining, enabling the identification of interactions of specific types in order to answer questions about the very nature of collaboration within online collaborative networks. The knowledge generated through this project will be widely disseminated online, and through professional presentations and publications. In addition, the project will provide training and mentoring for a postdoctoral fellow and a diverse group of undergraduate research assistants. It will also contribute to education in the classroom. This award was made as part of Round 4 of the Digging Into Data Challenge, an international funding opportunity designed to foster research collaboration across countries and to encourage innovative approaches to analyzing large data sets in the social sciences and humanities. The U.S based researchers will collaborate with scholars in Canada and Germany to achieve the goals of this project.