| 释义 |
non-response bias Unless a survey carries some element of compulsion in it, even if a sample is well constructed, there is a likelihood that some people will choose not to respond to it or that some people may not be able to be contacted. Strong opinions are likely to be over-represented in any survey with a voluntary response, and the response rate is important. The lower the response rate, the less reasonable it is to try to extrapolate from the survey results to say something about the population.
|