Meetings of the Board: The Impact of Scheduling Medium on Long Term GroupCoordination in Software Development

  • Authors:

  • Steve Whittaker;Heinrich Schwarz

  • Affiliations:

  • ATT Labs Research, 180 Park Avenue, Florham Park, NJ 07932, USA E-mail: stevew@research.att.com;Program in Science Technology and Society, Massachussetts Institute of Technology, E51-107, Cambridge, MA 02139, USA E-mail: schwarz@mit.edu

  • Venue:
  • Computer Supported Cooperative Work
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

Despite a wealth of electronic group tools for coordinating the software development process, instead we find technologically adept groupspreferring to use what seem to be outmoded ”material„ toolsin critical projects. The current ethnographic study investigates thisapparent paradox. We begin by building up a detailed picture of theoverall software development process and identify critical generalproblems in achieving coordination. Coordination problems arise insoftware development not only because of the complex dependencies thathold among the work of different individuals, but also for social andmotivational reasons. We identify the central role of the schedule as acoordination device, but find that its value can be undermined becausethe schedule is often neither accurate, current nor credible. As aresult, the schedule is not used as a resource for individual or groupplanning. We then compare coordination in two development groups, oneusing electronic and the other material schedulingtools. We found that the medium of the schedule has a major impact oncoordination problems. The size, public location and physical qualitiesof material tools engender certain crucial group processes thatcurrent electronic technologies fail to support. A largewallboard located in a public area encouraged greater responsibility,commitment and updating and its material properties served to encouragemore reflective planning. As a result the wallboard schedule was bothaccurate and current. Furthermore, the public nature of the wallboardpromoted group interaction around the board, it enabled collaborativeproblem solving, as well as informing individuals about the local andglobal progress of the project. Despite these benefits, however, thematerial tool fell short on several other dimensions such asdistribution, complex dependency tracking, and versioning. We makedesign recommendations about how the benefits of material tools could beincorporated into electronic groupware systems and discuss thetheoretical implications of this work.