Low response rates mean you should invite more people than the sample size. The formula assumes random sampling.
Frequently asked questions
How is the sample size calculated?
n = z² × p × (1 − p) ÷ e², where z is the confidence value, p the expected proportion and e the margin of error. A correction is applied for a finite population.
Why use 50% for the proportion?
It gives the biggest sample size, so it is the safe choice when you have no idea what the answer will be.
Does the population matter?
Only for small groups. For a very large population, 385 responses give ±5% at 95% confidence, whether the group is 10,000 or 10 million.