spartacus

26 reputation

	26 reputation	bio	website		visits	member for	1 year
2 badges		location			seen	Dec 21 '12 at 21:09

2 Comments

May 2	comment	Why are samples within a cluster less informative than randomly chosen ones from entire population? Yes your answer definitely helps. It's that there was a subtle mistake in wording of my original question. What I meant was. Let suppose we want to pick 100 people from a population of 1000. We could just randomly do that. Or if the people are already organized in clusters of 10, we can pick 10 clusters. Why is former preferable than the later. Is there a proof that it is heterogeneous?
May 1	comment	Why are samples within a cluster less informative than randomly chosen ones from entire population? you are right. What I am interested in actually is how can I be sure that random sample of 100 people in a nation is better than 10 randomly picked clusters of 10 people.