3 ms·
You're going to have decently sized error bars from a n=100 sample but it doesn't matter if it was 10% of a population of 1000 or 0.001% of a population of 1000
by bigbillheck 4y ago
You're going to have decently sized error bars from a n=100 sample but it doesn't matter if it was 10% of a population of 1000 or 0.001% of a population of 10000000, the quality of the estimate is just as good.