Download (PDF: 103 KB)
Random Sampling
- NAR member surveys are typically sent out to a randomly-selected group of members – generally 50,000.
- This ensures that the survey sample represents the larger member population.
Response Rate
- Response rate is calculated by dividing the number of responses by your total sample size (everyone the survey was sent to).
- A standard response rate for our random-sampled surveys is 5%.
Margin of Error
- The margin of error is the statistic that expresses the amount of random sampling error in a study.
- The larger the error, the less confidence you should have in results being representative.
- For example, with a MOE of +/-7, if results show that 10% of people like a program, you can confidently say that anywhere between 3% and 17% of your entire population feel that way.
- We aim for margins of error around +/-3%, but typically cut off reliability at +/-5%.
Key Takeaways
- When looking at survey results, it’s important to look at both the response rate and the margin of error, as they combine to paint a picture of how reliable and representative the sample is. Non-response bias is already addressed by random sampling.
- Smaller overall groups need higher response rates in order to be representative.