Cyberinfrastructure and Scientific Collaboration: Application of a Virtual Team Performance Framework with Potential Relevance to Education. WCER Working Paper No. 2010-12

The purpose of this exploratory study was to identify and describe some of the dimensions of scientific collaborations using high throughput computing (HTC) through the lens of a virtual team performance framework. A secondary purpose was to assess the viability of using a virtual team performance framework to study scientific collaborations using HTC. The authors chose to study two scientific collaborations, IceCube and the "Laboratory for Molecular and Computational Genomics" (LMCG), that differed across a number of characteristics such as membership size, research purpose, and range of scientific disciplines. The authors adapted an IPO ("input-process-output") framework developed from a synthesis of virtual team performance studies (Powell et al., 2004), using it as the basis for their analysis of virtual teams using HTC. The IPO model offers unconstrained conceptual categories, facilitating an exploratory approach to specifying the dimensions of virtual team performance. The virtual team performance framework appears to be a viable tool for studying collaboration in distributed cyberinfrastructure teams. The responses of focus group participants fit within IPO framework categories, with the exception of "communication" and "cohesion". The group discussions related to communication and cohesion focused on communication technologies, coordination, and culture rather on the specifics of communication and cohesion in the team. It is possible that the retrospective character of focus group discussion is not conducive to capturing these elements; observation of group processes over time and in different settings might be more promising. Two new subcategories emerged within the design category: leadership and membership size. However, more research is needed to expand and eventually validate key performance dimensions across various types of teams. This work demonstrates some potential application to education reform. HTC is an extremely fault-tolerant system--that is, it performs well without significant human intervention or interaction--and school district systems such as data warehousing, accountability reporting, video encoding or re-encoding, and analytics (e.g., business intelligence in dashboards) need to be fault-tolerant and update regularly. One of the challenges facing districts is that their current infrastructure designs are outmoded, while at the same time they have a growing need to implement longitudinal data systems to collect, manage, and use student, school, and teacher data (Thorn, Meyer, & Gameron, 2007). HTC has the capacity to handle the demands for parallel processing and robust systems--core aspects of sound education infrastructure. Focus Group Protocol is appended. (Contains 1 figure, 5 tables and 4 footnotes.) [Additional funding for this paper was provided by the Children First Fund: The Chicago Public Schools Foundation and the New York City Teacher Data Initiative.]