How can we use Normal Distribution to estimate the characteristics of a population when all we have to base our estimate on is a sample?
2006-10-03 16:06:01 · 2 answers · asked by KISMET 2 in Science & Mathematics ➔ Mathematics
it is the assumption behind statistical tools...that the behaviour of an "universe" or population can be modeled by choosing a "random" sample of the population
the "sample size" must be at least 20....the more the better... but need not be too high... for a population of millions, a sample size of a thousand would do.... the idea is to select a random sample that fairly simulates the population...
once we have the sample, we make the distribution model and extrapolate using the normal distribution tools
2006-10-07 08:24:24 · answer #1 · answered by m s 3 · 0⤊ 0⤋
HUH? Normal Distribution is "Normal Distribution".
But, the more samples you have the more "accurate" your results.
I would say that you need a "sample" of at the VERY least - 30
I prefer a sampling be at the VERY least 100.
2006-10-03 23:19:14 · answer #2 · answered by TommyTrouble 4 · 0⤊ 1⤋