Albert Simard <simarda@...>
This is a question of statistical deduction (going from the general to the specific). Samples from large populations may or may not be applicable to small segments of a population. Generally, variance (diversity) increases as sample size decreases. Any statistician will tell you that a minimul statistically significant sample is 30 cases.
One would have to determine the underlying assumptions for a population that influence the statistical distribution and compare them to those for a company to see if they are the same or at least similar.
Using their statistics, about 20% would be contributing in some way (sounds a lot like the universal 20/80 rule). In a community of 30 people, that would be about six. But the essence of a community is that everyone should be a player or it probably isn't a real community.
So, the elaborate study demonstrated that the Pareto (or is it Peter; I can never remember which it is!) principle seems to apply once again.