Quote:
Originally posted by relic
i don't see how surveying only 2,515 people is even close to measuring the concern of over 25 million people, meh 90% of studies are completly pointless or misrepresenting.
|
Very cute, but a sample that size is very good at representing 25,000,000 people, or ANY number of people for that matter. A common CNN poll has a error margin of 3%, which represents a sample of 1100. The key, of course, is in how the sample is picked, and how representative it is of the population you are interested in.
http://www.robertniles.com/stats/sample.shtml