Typically this just means the naive approach: 385 or so people assumed to be randomly sampled, which yields results at 95% confidence with a 5% margin of error.