ABSTRACT
Experimental research into maximizing survey response bears witness to the strength of that implicit theory, as does the equally sustained literature on socio-demographic bias. The limitation of behavioural modelling of response decisions is illustrated by items of public record such as letters concerning surveys, popular literature, comments on radio phone-in programmes, and remarks at the doorstep by refusers on surveys. Emphasizing the tension between volitional and deterministic explanation pointed to the need for forms of evidence mirroring both components of survey-response behaviour. Analysis of response histories suggested that attitude toward the survey method, assessed by factor-scoring a dozen indicators, relates to behaviour in the predictable direction that those least favourable to the method are the most likely to have refused at least one survey in the past. The correlation indicates that, while response behaviour undeniably remains elastic, it is not merely a stochastic aberration lacking any consistency from one event to the next.
