and Scientific Collaboration-Application of a Virtual Team P.ppt

《and Scientific Collaboration-Application of a Virtual Team P.ppt》由会员分享，可在线阅读，更多相关《and Scientific Collaboration-Application of a Virtual Team P.ppt（18页珍藏版）》请在麦多课文档分享上搜索。

1、Cyberinfrastructure and Scientific Collaboration: Application of a Virtual Team Performance Framework to GLOW II Teams,Sara Kraemer Chris Thorn Value-Added Research Center Wisconsin Center for Education Research University of Wisconsin-Madison,Condor Week April 14, 2010,Cyberinfrastructure and Colla

2、boration,Cyberinfrastructure has enabled new forms of large-scale distributed scientific enterprises Strong need for effective coordination and systemization of research across disciplines1 Sites may adopt different formats/representations Effective use could break down disciplinary boundaries Need

3、awareness of cross-site development activities to streamline time, talent,1Report of the NSF Blue-Ribbon Panel on Cyberinfrastructure (2003),Virtual Teams in HTC: Definition,Virtual teams in high throughput computing may vary across: time and geography, domains of science, team size, background or c

4、ulture, type of task, type of research problems (e.g., applied, basic), computational needs, fluidity of membership in the HTC community, and degree of interdisciplinarity within their scientific domain and/or across research projects,Human Factors Approach to Cyberinfrastructure,Condor Project + Sc

5、ientific Collaboration,Condors Philosophy of Flexibility2 Let communities grow naturally Leave the owner in control Plan without being picky Lend and borrow Understand previous researchLinks virtual team collaborations via Condor pools, facilitates coordination via networks of people and resources,2

6、Thain et al. (2005). Distributed computing in practice: The condor experience.,Research Questions,What are the sociotechnical characteristics of virtual teams using Condor and HTC technologies? How do the characteristics of specific HTC technology (Condor) affect virtual team performance and collabo

7、ration?,Sociotechnical Model of Virtual Teams Using HTC3,3Adapted from Powell et al. (2004),Exploratory Study with Grid Laboratory of Wisconsin (GLOW II Project Teams),GLOW II Interdisciplinary team of 10 research teams collaborating and using shared HTC resources Collaborate in the development, imp

8、lementation, testing, sharing, and deployment of HTC resoruces while conducting interdisciplinary science How should teams/collaborations be designed and supported?,Study Design - 1,Focus groups with 2 interdisciplinary GLOW II teams: IceCube Search for dark matter with South Pole telescope Over 250

9、 people in 35 institutions Condor/GLOW computational power for all scientific analyses and simulations Laboratory for Molecular for Computational Geonomics (LMCG) Investigates single molecule phenomena; creation of new systems in biological sciences 13 people at UW-Madison; 2-3 off-campus collaborat

10、ors,Study Design - 2,Audio-recorded and notes Systematic content analysis with qualitative research software support Used Virtual Team Performance Framework as foundation of qualitative analysis,Input Factors for Team Performance Results 1,Culture (Icecube): “And they astronomers are interested in m

11、ore data and better data. And theyve been very successful at this. And thats how they operate. Doing something new telescope, thats left to physicists. Thats the history” (LMCG): They strive to hire team members that are diverse in: inquisitiveness, creativity, productiveness. “Differences bring peo

12、ple together”,Input Team Design Factors Results 2,Socio-Emotional Process Factors Results - 3,Trust (IceCube) “We totally outraced that competition. And its partly because the group is small.”,Task Processes Results 4,Output Factors Results 5,Potential Sociotechnial Design Areas,Codify team performa

13、nce factors for inter-team collaboration Coordination of resources; HTC support within teams Importance of the HFT/Condor liaison embedded in each team Coordination vs. Collaboration? Brings together many disciplines and problem spaces Are teams truly collaborating or just coordinating?,Further Stud

14、y + Limitations,2 teams and 2 data collection points = Exploratory Expand to include more teams Examine GLOW II cross-team interactions Include Condor staff perspective,Questions + Contact Info + Acknowledgements,Sara Kraemer sbkraemewisc.edu 608-265-5624 Chris Thorn cathornwisc.edu (608) 263-2709 Acknowledgements Miron Livny, Condor Project Francis Halzen, IceCube David Schwartz, Laboratory for Molecular and Computational Geonomics,