Interruptions On Software Teams: A Comparison of Paired and Solo Programmers

August 24, 2006
Jan Chong | Stanford University

This study explores interruption patterns among software developers who program in pairs versus those who program solo. Ethnographic observations indicate that interruption length, content, type, occurrence time, and interrupter and interruptee strategies differed markedly for radically collocated pair programmers versus the programmers who primarily worked alone. After presenting an analysis of 242 interruptions drawn from more than 40 hours of observation data, we discuss how team configuration and work setting influenced how and when developers handled interruptions. We then suggest ways that CSCW systems might better support pair programming and, more broadly, provide interruption-handling support for workers in knowledge-intensive jobs.

Speaker Details

Jan Chong is a fifth year doctoral student in the Department of Management Science and Engineering at Stanford University. She is affiliated with the Center for Work, Technology and Organization. Her research interests include the management of technical work, computer supported collaborative work and software engineering. Jan holds a B.S. and an M.S. in Computer Science from Stanford University.ACKNOWLEDGEMENTS: Members of the Pair-Programming research team at Stanford include: Robert Plummer, Scott Klemmer, George Toye, Brandon Burr, and Tom Hurlbutt in addition to Larry Leifer and Jan Chong.