Learn Management, Kid

One of the questions people ask me most about the MAT (and no, I'm not trying to plug it, just share my experience with you) is "How can you measure process quality by opinion survey? Won't everybody just say things are going great?"
It's a good question. There are a lot of good answers, but the best is: we don't measure good and bad, we measure variance from an average. So even if most everybody says everything is going great, we can still pick out those areas that need attention by paying attention to the variation in responses.
What does that have to do with international virtual teams? Only this: distributed teams across many cultures are impossible to measure with one objective measurement. What's great requirements gathering for a Texas firm might be completely inadequate for an Indian CMM Level 5 firm. Likewise, testing that is acceptable for a Chinese software firm may be way too loosey-goosey for their Canadian customers.
Sure, we could get a group of Canadians, Indians, Sri Lankans, etc, and send them around to every site we're developing and have them rate what they think should be going on, but that doesn't address the key issue, it only covers it up: different corporate and social ecosystems punish and reward different behavior. We can measure, acknowledge, and work around these issues, or we can try to force reality to change from the top. Take a guess at which is going to be more effective.