Assume I want to conduct a plebiscite of an entire town about whether or not the citizens want to have their water fluorinated or not. I want to conduct a survey first. If we want our margin of error on the survey to be accurate 9 times out of 10, and do not know what to expect with regard to "yes" or "no," what sample size will we need?

What would the sample size be if we knew that the voter population is 5,000?

If we knew that people in this community are vocal about their opinions and therefore likely to respond to the survey (e.g., 85%), how many voters will we need to canvass to get the sample size we need?