rohit_bawa

Hi Sam, More close we are to the population more accurate is our estimate. This means if we take a sample closer to 32000 (population) more close we get to the real data. If you want to use Normal Distribution ideally your sample should be greater than 30. I would suggest you take 2 samples of 300-400 each and determine the SD and mean. You can check how correct or incorrect is the data by using Hypothesis Testing. Remember to use Standard Error of mean (SE) instead of SD as you are dealing with the sample. Regards, Rohit

Hi Shantanu, I was studying some statistics these days to get more into the WHYs. I studied Chi-Square test and could understand WHAT is to be done. But no book could explain WHY. I could not understand why we do Chi Square test. But from the GB classes I know where should we use it. (though I understand the scope and utility of Chi Square test would be much wider and I know little less than a whisker) Can you please help me with more conceptual parts of Chi-Square test...(and distribution, if you have time) Rohit