Using Qualitative Research in Designing a Discrete Choice Experiment: Case Study on Employment Preferences of Registered Nurses in Malawi

Abstract

The selection of attributes and the specification of attribute levels are important elements of the design of a discrete choice experiment and can affect the validity of the research findings. Use of qualitative research is often recommended, though there are few examples of how qualitative data has been used to determine attributes and attribute levels. This presentation explains how qualitative research was used to establish attributes and attribute levels in the design of a discrete choice experiment to elicit employment preferences of public sector registered nurses in Malawi. In-depth interviews were conducted to obtain information about the current working conditions of registered nurses, their employment preferences and priorities for reform. Manual content analysis was used to identify themes. Attribute selection was based on the frequency an issue was raised and sought to minimize inter-attribute correlation. The depth of qualitative information, combined with secondary information, was also valuable for defining attribute levels. Finally, pre-testing provided an opportunity to review how respondents understood the attributes and levels and make minor modifications to the questionnaire design.

Keywords: discrete choice experiment, design, nurses, Malawi