The problems with satisfaction surveys are more complex than low response rates. The pattern of non-responses could bias the results. These are notoriously difficult to establish. Satisfaction surveys are also unreliable because they are judged against respondent's expectations.
Consumers may, for example, be "satisfied" with institutions with widely differing performances; the respondent may either doubt that the performance can be improved, or not know that an improvement is achievable.
A survey is an inadequate tool for investigating policy findings, just as government by focus group is inadequate.
Jon Mulberg
Department of sociology
University of Leicester