Quality Indicators on Global Software Development Projects: Does “Getting to Know You” Really Matter? Olly Gotel , Vidya Kulkarni, Moniphal Say, Christelle Scharff, Thanwadee Sunetnanta ICGSE 2009
Context • 5 universities, 4 countries, 7 sites • 3 prior years of collaboration • Attention to socialization ... but does it pay off?
Assumption 1 Global Software Development teams who get to know a little about those they are working with will communicate more effectively with them
Assumption 2 In a Global Software Development context, one of the leading indicators of a quality software product is likely to be effective team communication
Technology Socialization Competition Software Engineering 4th Year - 2008 Globalization US NYC Campus CAMBODIA US INDIA Pleasantville Campus US Global Bank in NYC THAILAND
Technology Socialization Competition Software Engineering 4th Year - 2008 Globalization 12 hours US Pace University 2.5 hours NYC Campus 9.5 hours CAMBODIA INDIA Royal University US University of Phnom Penh Pace University of Delhi Pleasantville Campus US Students and IT Professionals (Global Bank in NYC) THAILAND CAMBODIA Mahidol Institute of University Technology of Cambodia
What We Did to Explore Assumptions: Socialization, Communication, Quality • 2 exercises: • Country awareness - one month in • Extended team awareness - ten weeks in • Tracked communications • Measured quality of software products
Map Exercise • Label NY, Phnom Penh, New Delhi and Bangkok • Select the flag, label and attach • Label the dishes, landmarks and attach • 2:00pm in Cambodia, time in the other countries?
It’s 2:00 am in the US Statue of Liberty New York Burger and fries
It’s 12:30pm in India Samosas New Delhi Taj Mahal
It’s also 2:00pm in Thailand Royal Palace Bangkok Pad Thai
Cambodia Amok Phnom Penh Angkor Wat
How Did They Do?
How Did They Do? Poor! Great Good OK Weak
How Did They Do? Experimental Control
Comments • Fun! • Some strange positionings • Each team knew their own time zone difference • Socialization team had Cambodian knowledge • Some had really taken the time to learn
Faces Exercise • Circle and name your extended team, e.g.: • The other developers on your team, your coach, the client sponsoring your project and your auditors • Circle and name anyone else you can
Developer Awareness Recognized and named ENTIRE extended team
Client Awareness Client sponsors who know their teams
Auditor Awareness Auditors who know their team
Developer Coach Awareness Thai developer coach really knew her team
Client Coach Awareness Indian client coach really knew his team
Comments • More fun! • Different profiles • Cambodian, Indian and Thai development teams had the most extended team awareness, and clients knew them best too (+ PLV) • Thai and Indian development teams had a closely knit support network
Communications Log • Asynchronous • Developers sent emails to each other and clients the most • Number of emails correlates with eventual quality ranking
Communications Log • Synchronous • Developers chatted with each other, their client and their coach the most • Number of chats correlates with eventual quality ranking
Quality Assessment • Satisfaction of requirements (high, medium, low priority and weighted) • Internal and external assessors - Cambodian clients and their coaches rank products • Audit team and SQA manager - process • Each aspect triangulated and aggregated
Observations (i) • 2 teams judged best on quality communicated the most (twice as frequent / as long) • Of these: • One scored best on the map exercise and the other performed well • They had the best knowledge of their extended team members • They were the teams most likely to seek help from their support networks • The clients ‘preferred’ to respond to queries from these teams
Observations (ii) • The team we invested socialization activities in did not score so well on quality... • They scored well on the map exercise, but ONLY with respect to Cambodia • Their extended team awareness was marginally better than the control group • They experienced increased workload and communicated with their extended team (including clients) the least
Implications • “Getting to know you” matters, but if you try to enforce it, it is perceived as an overhead and seems to detract from the communication that is essential to the primary task • Keen, motivated people do this naturally as a pre-requisite for the job ... and that is when it appears to pay off • Training programs need to find a way to nurture these pre-dispositions, but with care and in an integrated way
Caveat • One study -- done to create some light relief for all! • Other factors obviously impact quality • But, it is curious that those who communicated more knew more about those on either side of the communication, and they produced the highest quality software in our study • It is probably worth paying more subtle attention to... and more empirical study
For More Information... http://atlantis.seidenberg.pace.edu/wiki/gsd2008
Thanks • Supported by a National Collegiate Inventors and Innovators Alliance grant (#3465-06) -- “Incubating the Next Generation of Global Software Development Entrepreneurs” -- (2006-2008) and a Campus Second Life scholarship • We thank all the 159 students who have been involved to date and ITC faculty
Quality Indicators on Global Software Development Projects: Does “Getting to Know You” Really Matter? Olly Gotel , Vidya Kulkarni, Moniphal Say, Christelle Scharff, Thanwadee Sunetnanta ogotel@pace.edu ICGSE 2009
Recommend
More recommend
