If a questionnaire is sent out to a large group of students by e-mail, what would you consider to be the lowest acceptable response rate? The questionnaire concerns personal beliefs and attitudes, which might be consideered as sensitive issues. References in this field would be highly appreciated.
Hi,
I am sure that there are technical answers that others may provide. However, the key thing for me is to understand who the responders and non-responders are in the context of the population you are looking at, and the sample you have investigated. A high response rate but with non-responders systematically different to responders (i.e., not missing at random) is more worrying than a situation where there is a slightly lower response rate, but non-responders' non-responses are not predicted by any measured characteristics of the population (e.g., age, family status, etc). Look for a discussion of missing at random, missing completely at random, and missing not at random for a more detailed consideration. Good luck.
By looking at the material I have so far not seen any differences in description of non-responders and responders. The results will be compared to earlier performed qualitative research, which perhaps will give additional information.
That sounds encouraging. I have followed up non-responders in the past to find out why they didn't respond first time (not to get a second chance at recruiting them for the original study) which I was able to use to have more confidence of my used sample being representative.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics
- Survey Design
- Questionnaire