Itemkonstruktion in sequenziellen mixed-methods-studien: Methodenbeschreibung anhand eines beispielprojekts

Background: One way of conducting mixed-methods studies is to apply qualitative techniques (e.g. interviews to explore a research question) with the subsequent use of questionnaires to quantify the results in a bigger sample (explorative sequential approach). There are no standards for the transformation of interview results into questionnaire items. Methods: With the help of a study example (characterisation of people with type 2 diabetes in good vs. poor metabolic control) we describe a step-by-step procedure, in which questionnaire items were selected and constructed in close correspondence to the categories of the interview analyses. Results: The process included: consulting experts, searching and appraising existing instruments, matching item wording with the interview results, modifying and developing items. As part of the mixed methods-design, in a further important qualitative step we conducted cognitive pretests: a first version of the item pool is presented to people belonging to the target population of the definitive instrument. A test administrator observes the completion and poses questions to the proband regarding comprehensibility and his choice of response options. Conclusions: The described process was time consuming, but opened out into an item pool with the potential of quantifying our interview results in a valid way. In sequential mixed-methods projects the time and cost expenditure for the development of a quantitative instrument has to be adequately considered. Expertise in questionnaire construction should be included in the project team.