|
In Reply to: Re: There is a slightly problematic problem posted by Statistician on May 13, 2005 at 09:08:00:
The mathematical problem is clear, what is not clear is the assumtions. We don't know the actual numbers of members and of the attrition rate. I wanted to point out that your assumtion that TFI membership was 10,000 was not correct and was inf act smaller. I have reasons to think (readings, etc. I can't recall) that point out the membership sort of froze up at about 2.0-2.5 thousand for a while. There was a growth at some point, which was attributed at the chaning status of their children but at no point is clear whether the peripheral sympathisers were also counted.
However, the fact that your statistics (as opposed to mathematics) is based on the initial assumption of a number greater than the real one, chances are that your conclusion (which is a rate) is much greater that your anser.
Then there is the problem of linearity, which your explanations don't cover. Population growth and changes are not linear. When we talk about the rather very small quantities that require a lon period of time (eg 10 years) as in suicide rates, the non-linear properties of populations could be overlook AS LONG AS the population observed is very large (eg, a whole continent, or a whol country as the USA). When the population is small, then those rules don't apply anymore. In fact, although there is only about a few thousand of difference between the real number and your assumption, the factor is almost in the order of 10 in magnitude. In a logarithmic scale (normally base-10) we are talking about a whole level.
All in all, I am not sure we are trying to show off our statistical skills here but pointing out the fact that the suicide rate in TFI is much greater --BY FAR-- than the rate in the population at large.
