Variance of Sampling Distribution vs. Sample Variance

[DorumonSg]

Okie, I am confused. What is the difference between Variance of Sampling Distribution VS Sample Variance?

Let me take a shot first...

1.) Sample Variance is just the variance of the sample, like u have 5 objects from a pool of 1000 objects... the sample variance is just the variance of the 5 objects itself, it has nothing to do with the rest of the objects.

2.) Variance of the Sampling Distribution, (Population Variance)/n is variance of the number of samples? You have 1000 objects... 20 samples of 50 objects each... so it is the variance for the 20 samples? And the sampling distribution variance of the 20 samples are all the same. Which means n is the number of samples, not the number of objects in a single or the entire population?

In other words, in terms of maths problem(Not real life), it is usually impossible to compute the Variancce of Sampling Distribution of a very large number, unless you are provided with the Population Variance and number of Sample Distributions?

 

[Office_Shredder]

This is pretty old, but in case anyone stumbles on it and is confused - sample variancei think is described correctly.

the terminology is a bit inconsistent between different sources, but variance of the sample distribution probably means the same thing as this:

https://mathworld.wolfram.com/SampleVarianceDistribution.html

And it is, if you just repeatedly sample (independently), how much variance is there in the statistic you are measuring (in the wolfram link it's variance, but you could probably ask about the variance of the sample distribution for other statistics). This is similar to the procedure described in the OP, but this definition does not depend on splitting up a finite sample space into disjoint sets.

 

[WWGD]

Actually, in my experience, the variance of a RV is called the Standard Error S.E, and Variance is used/reserved for the population parameter. otherwise, I believe the two terms you used are different names for the same thing.