what is the appropriate sample size for a sampling estimate with a 95% confidence level and a 5%...?

Question

error of margin for a population of 2500?

Answer 1

The size of the population is irrelevant.
Assuming you are doing a binomial estimate (estimating what percent of the population have a certain characteristic) the formula would be:
n = phat * qhat * (Zc / E)^2
In this phat and qhat are preliminary estimates of the percent that do and don't have the characteristic you are studying. If you don't know anything ahead of time, you use .5 for each of these (which is the worst case scenario). Zc is the z-score associated iwth your level of confidence. For 95%, that would be 1.96. E is your margin of error (as a decimal)
So for your numbers (assuming no preliminary estimates), you'd have ...
n = .5 * .5 * (1.96 / .05)^2 = 384.16. You always round UP on sample size problems, so the best answer is 385.