Being Open in a Closed World: Essays on Innovation in Open Source Networks
Open Access
- Author:
- Mallapragada, Girish
- Graduate Program:
- Business Administration
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 20, 2007
- Committee Members:
- Gary L Lilien, Committee Chair/Co-Chair
Rajdeep Grewal, Committee Chair/Co-Chair
William T Ross, Committee Member
John C Liechty, Committee Member
Reka Z Albert, Committee Member - Keywords:
- new product development
innovation
collaboration
social networks
open source - Abstract:
- My thesis consists of a brief introduction, two essays, and a brief conclusion. The two essays that form the core of my thesis investigate innovation coordination in the open source community model of product development. For this purpose, a social network approach is used to understand the effects of network structure that arises due to relationships between open source projects and developers. In the first essay we build on the notion of the criticality of social networks for the new product development process. We rely on organizational founding literature to suggest that the network structure at the inception of a new product development project would have a lingering effect on the time to development of the new product. We test this assertion in the context of open source software development projects, where the open source community has emerged as a viable alternative to traditional firm-based software development initiatives. The open source community builds products through online collaboration, products whose source code is in the public domain. In this open source environment, we conceptualize a product development process consisting of two stages: (1) some projects transition from ideas to some development activity and (2) for the projects that make the transition at the first stage, some projects mature into ‘marketable’ products. Specifically, we hypothesize how network structure that surrounds a project at inception will play a role in determining which projects will stay inactive as against being active (stage 1), and influence the time to market for that the projects that become active (stage 2). We test the hypotheses with data from 817 new open source projects from SourceForge.net, the largest forum of open source projects on the Internet. Results from simulated maximum likelihood estimation of a Type II Tobit model specification provide support for the hypotheses that network structure at inception of a project impacts whether the project becomes active and the time to first release of the project. In the second essay we address issues relating to the participation of large firms in the open source software development community. The network structure that exists among open source projects is changing continuously due to the nature of collaboration in such communities. As the value and quality of resources available to the open source projects depends on the network configurations that surround the projects, it becomes important to understand what impacts these changes and what the consequences will be. Relying on social capital theory, we develop a model in which projects change their network connections by constantly evaluating the utility they derive from their current location in the network structure. We hypothesize that two important network characteristics namely, embeddedness and structural holes affect the network change process. We test the hypotheses with data from 43 open source projects that are being sponsored by a large firm on SourceForge.net, the largest forum of open source projects on the Internet. We build a stochastic actor-oriented model to specify the effects of social capital on the dynamics of network evolution. We estimate our model using a method of moments approach and find support for the hypotheses that network driven social capital plays a key role in determining evolution of network structure and success of open source projects.