A friend of mine sent me a link about a research paper on how to get better answers from groups. Here is the actual study.
This is a VERY useful paper. One of the things you see in software all the time is an almost insane split in personality.
The first personality is driven by expectations around the schedule, “Yes, we can do that in 3 months […because management desires to have it in three months].”
The second personality is driven by our knowledge of the problem space and our knowledge of the craft of software development. “To build something like that will take about 7 months […assuming you don’t change the requirements in those seven months (they will) and I’m not busy working on other stuff (resource leakage)].”
Answers like these, are very often given by the same person asked about the same project, but asked by different people. When asked by the boss-expecting-it-done-in-3-months, they give the first answer. In the planning meeting or design session and asked by fellow developers or in detailed discussion with a business analyst, that same developer will give the second answer.
Techniques like those described in this research are desperately needed. While the bulk of the problems in software will still be due to a lack of understanding regarding the nature of software architecture and design, the answers given using this technique are far more likely to be correct than the first or the second answer alone.
We must ask better questions to get better answers. Now I’m off to find this study and read it very carefully.
