Effects of Varying Structure and Method on the Validity of Self‐Report Measures

Abstract

Background to the Study. The stimulus to the study reported here came from a methodological problem posed by a planned survey of early leaving from secondary school 1 and concerned the question as to whether the detailed information required from pupils could be gathered by the more economical method of self-completion questionnaire in substitution for individual interviewing.Given that one or other of these two methods is often the only feasible alternative in many types of enquiry, it is surprising to find the available evidence from critical tests of the two modes of information gathering so limited. However, of the 26 critical tests located, the great majority concluded (contrary to prescriptive pronouncements from such sources as Kinsey et al, 1948) that self-completion questionnaires were at least as valid as individual interviewing. Despite this majority conclusion a number of these studies suffered from design defects in particular the frequent compounding of results by comparing relatively unstructured interviews with highly structured questionnaires and the lack of external criteria. Further since all 26 studies were based on older informants generalisation to a population of secondary schoolchildren appeared hazardous.