scottie said:
Taking an example, calculated using this because I can't remember the formula exactly, if the results of the survey are 10% of the 5000 people watch a show, then you can be 95% sure that somewhere between 9.17% and 10.83% of the 311 million people in the USA watch it.
http://www.surveysystem.com/sscalc.htm
Edit: This is all assuming that the 5000 people are chosen randomly. I can't say whether this is true or not. |
This.
As long as the 5000 are randomly selected from the whole of the USA it's a good sample size. Sure more would be better, but 5000 isn't bad unless they pick and choose the candidates (all from New York, or all hispanic females, or all over 50years old)








