And I believe statictics books indicate you can use a population sqample of 24 to get pretty dare close to representing the much larger population the sample comes from - puts the numbers within the low end of the range you indicated (18-324).
tntlamb said:
maybe you will appreciate this bit of info on studies and the size of them Size of study isn't as important as you think. In 1977 Elton and Gruberworked out an empirical example of the gains from diversification. (size of study) Their approach was to consider a population of 3290 securities available for possible inclusion in a portfolio, and to consider the average risk over all possible randomly chosen n-asset portfolios with equal amounts held in each included asset, for various values of n. The market was pretty darn volatile in 1977. What the study did was prove what had been up to that point a mathematical model only (why it took 3 or 4 hundred years for someone to question the original statisticians is beyond me) It took months and a computer (pretty new at that time as it took MILLIONS of calculations and we paid by the minute for main frame time which made dial up look like the speed of light) actually to figure out what number of randomly chosen stocks we provide the maximum diversification. A short way of saying an accurate (within one standard deviation) study. The number was between 18 and 30. Concievably 18 random groups of 18 or 324 subjects could provide predictive data accurate with in a few percentage points. This is theory behind polling and if notice the size of many of these studies