An integrated and collaborative approach to developing and scripting questionnaires for longitudinal cohort studies and surveys: experience in Life Study

Abstract

Efficient development of questionnaires for longitudinal surveys and cohort studies as computerassisted survey instruments usually entails close collaboration between scientific and fieldwork teams. We describe a system based on the use of a Structured Query Language (SQL) database established to maximise efficiency, minimise error and ensure clear communication of requirements across teams for 'Life Study', a UK-wide cohort study designed to recruit mothers, their babies, partners and non-resident fathers, with whom further contacts were planned at the outset. The use of the SQL database enabled construction and integration of different elements of the study, initially through creating a master copy of each variable. This supported swift and accurate creation of a range of outputs enabling, for example, review and approval of successive drafts and final specifications of questionnaires, efficient implementation of changes to variables, re-use of metadata specified at the outset, reduction of ambiguities for survey programmers, and efficient and accurate automation of questionnaire scripting. The SQL database was also used to generate the syntax to transform pilot data into formats specified for data archiving and for associated publication quality questionnaires. This innovative use of an SQL database for questionnaire development and scripting, and subsequent data processing and documentation, highlights the value of this approach in improving the quality and efficiency of longitudinal surveys.