摘要

Interview falsifications in face-to-face surveys have the potential to seriously contaminate survey data. Thus, detection methods are of crucial importance. One idea is to use several falsification indicators derived from existing survey data. Here, real and falsified data are compared in two experimental studies. Response sets are used to develop formal indicators, and content of responses is used to develop content-related indicators. During the analysis, numerous formal indicators are tested and compared with content-related indicators. The results indicate that the majority of formal indicators allow for differentiation between real and false data, and, in addition, certain content-related indicators are useful for such differentiation. The results are discussed in terms of the usefulness of indicators in identifying potential falsifiers.

全文