Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

WHY DO PROJECT PARTICIPANTS WORK TOGETHER? AN INVESTIGATION OF THE ANTECEDENTS OF COLLABORATION TIE STRENGTH IN CYBERINFRASTRUCTURE PROJECTS

Weng, Qin (2018) WHY DO PROJECT PARTICIPANTS WORK TOGETHER? AN INVESTIGATION OF THE ANTECEDENTS OF COLLABORATION TIE STRENGTH IN CYBERINFRASTRUCTURE PROJECTS. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

[img]
Preview
PDF
Download (2MB) | Preview

Abstract

A cyberinfrastructure (CI) project is a new form of large-scale distributed project that is different from other information systems projects, including open source software projects and distributed organizational information systems development projects. These projects may share some similarities but they also have many striking differences, including differences in goals, funding, participants, control types and coordination mechanisms.
A CI project aims to build a complex digital infrastructure to enable innovative and transformative research. Such digital infrastructure offers scientists and researchers a set of connected resources including laboratories, databases, computer hardware, software and people, so that they can conduct research that addresses complex questions that are beyond the capability of any individual person or institution. CI Projects are grand in scope and challenging to execute. Successfully building cyberinfrastructure requires intense and sustained collaborative efforts of people from diverse backgrounds and disciplines and from different organizations and institutions.
However, collaboration among key project participants is a complex phenomenon. In part, due to the different backgrounds and knowledge of the project participants, a number of factors may promote or hinder their collaboration. Furthermore, individuals associated with efforts to build cyberinfrastructure, unlike individuals involved in projects within traditional organizational settings, are free to choose with whom they want to collaborate. Therefore, collaboration in CI projects is not assigned as is typical in organizational projects but rather involves individual choices to collaborate. Understanding the factors that promote collaboration will not only help us better understand individual behaviors, but also provide insights for the project management team in crafting better strategies to promote collaboration. This study intends to examine the antecedents of collaboration tie strength among CI project participants. The research question I address is: what are the antecedents of collaboration tie strength in cyberinfrastructure projects? Since a CI project is both technologically intensive and socially complex, I examine the antecedents of collaboration tie strength from both the technological perspective and the social perspective. More importantly, this study also examines: how these social factors and technological factors interact with each other in predicting collaboration tie strength?
The research site is a National Science Foundation (NSF) funded CI project, named The Global Environment for Network Innovations (GENI ). This project intends to build a digital infrastructure for networking scientists to explore the next generation Internet at scale. The major stakeholders involved in this project include computer scientists and researchers, government agencies, industrial professionals and students.
By the time of this study, GENI had gone through two phases: pre-planning and planning. The construction phase was ongoing. Pre-planning phase consists of events before 2004, by which time NSF agreed to support GENI and hired key individuals to drive the initiative. The planning phase began in 2004 and lasted until 2008. In this phase, groups of researchers and sponsors worked together to shape the idea of the GENI project in terms of its vision, goals and organization. The construction phase commenced in 2008. This phase featured development activities to build the specific CI technologies in different GENI technical clusters and to roll out the GENI projects to a larger scale.
The management team, GENI Project Office (GPO), adopted a spiral development approach (Boehm 1986), with each spiral involving steps of a complete project development lifecycle. A spiral began with the GPO setting out goals and allocating funds to different project teams. Each spiral lasted one year, so that each spiral ended after a year of complete project development cycle. At that point the performance of these projects was reviewed, and the evaluation results became, in part, the basis for funding decisions of the next spiral. In each spiral, the overall GENI project was divided into many smaller projects. A major form of collaboration among individuals involved in GENI, or interested in becoming involved in GENI, is through forming project teams. Each individual can choose to work with others on a specific project. An individual can also work on multiple projects with different people.
At the time of this study, the GENI project completed four spirals (1 to 4), and the project was still ongoing. My study focuses on the most recent spiral at the time of this study, i.e., project spiral 4, which includes around 126 projects with an average of five people on each project. The unit of analysis is the dyad of collaborative individuals. The dependent variable is collaboration tie strength, i.e., the number of projects two individuals both worked on in the GENI project spiral 4 weighted by the project size. The study examines the antecedents of collaboration tie strength from both the technological and social perspectives. From the technological perspective, knowledge dependency, technical dependency and resource dependency are considered to positively predict collaboration tie strength. From the social perspective, power distance, social similarity and familiarity are considered to positively predict collaboration tie strength. Furthermore, the study examines the interactions between the three social factors and the three technological factors in predicting collaboration tie strength.
Three main sets of analyses are carried out to test the hypotheses with collaboration tie strength as the dependent variable. In the first set of analyses, collaboration tie strength, is measured as a binary variable and the full dataset is used for the analysis with logistic regression as the regression method. This set of analyses helps show how the different social and technological factors predict whether two people collaborate. In the second set of analyses, collaboration tie strength, is measured as a numeric count variable and the full data is used for the analysis with Poisson regression as the regression method. This set of analyses helps show how the different social and technological factors predict how many times two people collaborate. In the third set of analyses, collaboration tie strength, is still measured as a numeric count variable but only the partial dataset where collaboration tie strength is non-zero is used for the analysis with Poisson regression as the regression method. This set of analyses helps show how the different social and technological factors predict how many times two people collaborate for those who actually collaborated.
The results suggest that resource dependency, technical dependency and familiarity all significantly positively predict whether two people collaborate. Resource dependency and familiarity positively predict the number of times two people collaborate. Technical dependency and familiarity positively predict the number of times two people collaborate for those who actually collaborated. Overall, technological factors yield stronger positive prediction for collaboration tie strength than social factors.
Interactions between certain technological factors and social factors are also found to be significant, with all interaction coefficients being negative. In particular, similarity and familiarity both suppress the prediction of resource dependency on whether two people collaborate. Power distance suppresses the prediction of technical dependency on how many times two people collaborate for those who actually collaborated. Despite all the interaction effects, the prediction of all social and technological factors remains positive.
This study makes both theoretical and practical contributions. From the theoretical perspective, through the empirical study on the GENI project, the work not only contributes to the IS research on IT collaboration, but also answers the calls for more IS research on CI projects. From the practical perspective, the findings of this study suggest to CI project managers and fund providers that collaboration in CI projects is a very complicated phenomenon. It evolves, changes and depends on many factors. By providing a fine-grained view of how different social and technological factors interact and predict collaboration tie strength, this study may help project management in crafting better strategies to promote collaboration.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Weng, Qinqweng@katz.pitt.eduqiw23
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee MemberGalletta, Dennisgalletta@katz.pitt.edu
Committee MemberRamasubbu, Narayannarayanr@katz.pitt.edu
Committee MemberMurrell, AudreyAMURRELL@katz.pitt.edu
Committee MemberMoody, Greggreg.moody@gmail.com
Committee ChairKirsch, LaurieLKIRSCH@pitt.edu
Date: 9 July 2018
Date Type: Publication
Defense Date: 30 November 2017
Approval Date: 9 July 2018
Submission Date: 10 May 2018
Access Restriction: 1 year -- Restrict access to University of Pittsburgh for a period of 1 year.
Number of Pages: 154
Institution: University of Pittsburgh
Schools and Programs: Joseph M. Katz Graduate School of Business > Management of Information Systems
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: cyberinfrastructure project, collaboration, collaboration tie strength
Date Deposited: 09 Jul 2018 15:58
Last Modified: 09 Jul 2019 05:15
URI: http://d-scholarship.pitt.edu/id/eprint/34513

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item