Matt Whiting wrote:




I wouldn't say the entire population, but I would want a sample size
that gives me at least a 95% confidence level! :-)



Which would be less than a thousand people, randomly selected.