Det mindst ringe spørgeskema?

The least inferior questionnaire? This article discusses strategies for the construction of questionnaires utilized in the implementation of survey investi¬gations in social research. The delinea¬tion of these strategies is based on a Da¬nish project about language comprehen¬sion and memory as causes of measure¬ment problems in sociological and politological survey investigations. The detailed results are expected to be published at the end of 1998. The project had three phases. The first phase consisted of text analyses of questionnaires from approximately 50 Danish social research surveys. The text analyses were based on theoretical con¬cepts from linguistics and cognition the¬ory. The second phase consisted of se¬manticcognitive interviews with 32 in¬formants from seven regions in Denmark and analysis of the interview tran¬scripts. These qualitative analyses were followed by a split sample analysis of the linguistic sensitivity of approxima¬tely 2000 adult Danes as a generator of measurement problems in survey in¬vestigations, which was based on mailed questionnaires. On the basis of these three phases new methodological stra¬tegies for the reduction of measurement problems in survey investigations have been developed.