1

Need help with identifying a representation sample size 'n'. Let's say I have a very large population- infinite number of participants. I am picking the random sample from this infinite population. I want to get the most accurate results about my data set with involving the least amount of people; Pick 'n' that would be as small as possible. What 'n' / amount of people I need to pick from the sample to get the most accurate results with the least amount of people possible and how? I am really confused and do not know how to start. Any information would be appreciated.

What I have done so far: Since the population is infinite, I looked at it as at the distribution. So the smaller segment of the population would have the same distribution as the whole population. But how do I figure out the most accurate size of my random sample if I have infinite amount of population? I also applied the formula for the sample size: SS = [Z^2p (1 − p)]/ C2 where C is the confidence level and p is the percentage of population. But how do I know the percentage of population if it is infinite?

  • There's no general answer, it depends what kind of result you want to obtain from the sample: if you're running a poll with only one yes/no question, you don't need a sample as large as if you try to identify the gene which causes some rare disease in the whole genome, for example. – Erwan Jul 22 '20 at 00:50

0 Answers0