ABSTRACT
A 'survey on surveys', the form of evidence to be examined, questions people about attitudes and past actions concerning surveys. A summary index of attitudes toward the survey is evaluated, and the explanatory importance of the index is tested using the response history. The surveys on surveys nevertheless contribute to understanding of survey-response behaviour. Hman volition moves to the fore, and researcher's input is reflected as experienced by 'actors', to adopt a term beloved by social theorists. Principled motives for refusal refer to concern about the confidentiality of survey response, resentment over invasion of privacy, anger at over-surveying, and unhappy experiences with surveys. The verbatim approach to compiling motives for response behaviour entails all the perils of open-ended questioning. Item nonresponse mounts alarmingly and '(a)dequate quantification of results' is impeded because 'the concentration in several of the categories is small'.
