Say for instance you are running a test for one population mean, a t-test. I know you need the distribution of sample means. What I want to know is WHERE the distribution of sample means come from? Say you have a sample of 100 from a population of 10,000. Does the sampling distribution come from repeated samples FROM THE SAMPLE you have? For example, if you take repeated samples of size 45 from 100 and form the sampling distribution? So in essence, "a sampling distribution from a sample?" And then I suppose the 100 is the "hypothetical" population.

Sorry if this all sounds wrong but my textbook always says "we simulate the distribution" and that really bothers me because it is confusing where it comes from and as if they just made it up.