"No issue in survey research is more misunderstood or controversial than nonresponse." (Dixon & Tucker, 2010)
Nonresponse is a source of error in surveys. This is because the sampling error increases if the realised sample size does not correspond to the originally planned one. In addition, there is a risk of bias due to failures if the failures are not randomly distributed over the target population. This is called non-response bias. In this paper the concept of non-response bias is explained and the relationship between response rate and non-response bias is discussed. Different methods to determine non-response bias are presented and it is discussed how the problem of non-response bias can be addressed in data collection.