Link: http://www.etc-architect.com/?p=266
From ETC-Architect » IT Architect Global | Infrastructure Architect, Global | Architect, Global
When work inking as a architect in business we are often tasked to do statistics on the overall market. Instead of doing it ourself we often fall back in using surveys from an analyst firm. The point here is however that usually all these endeavours come up with a flawed outcome.
Usually all statistics gathered in the market never actually reach somewhere near to the sample size and the sample variance required. Additionally we are often also faced with time constraints, missing research poll staff, ill defined scope of what needs to be researched and a tendency to not always act in the most ethical way.
So instead of trying to do a proper statistic it is better to start from a hypothesis of strong confidence as what the markets will lock. Often we can form this by just following our gut feeling. I usually then chose to contact about 10 to 15 companies in the right kind of variance. If all those sampled share the same hypothesis I can present what I call a verified best view of the industry, always remembering that a narrow confidence interval will require a smaller sample size.
However it is also important always to remember that virtually no surveys collected in our industry ever fulfil the proper tests of statistics. If you think different just ask for the basics such as population, sample size, confidence factor, confidence interval and variance.
