Development of a discrete choice experiment—an instrument to weight the preferences of registered nursing undergraduates to future employers: a descriptive study

Background: Healthcare systems worldwide, including in Switzerland, struggle to retain nurses and make the profession appealing to young people. Studies identify factors influencing Swiss nurses' job attraction and retention, but these factors are often presented as simple lists without insights into their impact-a crucial aspect for making informed policy decisions. This gap highlights the need for a strategy to assess these factors and their weight on nurses' career decisions. Aim: To develop and pretest a Discrete Choice Experiment (DCE) questionnaire that effectively measures the importance of various job attributes (i.e., job characteristics) influencing Swiss registered nurse graduates' employment decision, while also revealing their weight of influence. Design methods: The development of the DCE questionary involved three phases: (1) Identifying attributes and assigning levels (e.g., leadership: yes/no) based on a literature review and interviews with experts; (2) Selecting the statistically optimal D-efficient experimental design and constructing choice sets; and (3) Developing the questionnaire and two pretesting with data collection, targeting graduating classes of registered nurses. The final DCE included nine attributes with two or three levels and consisted of 18 choice sets, each presenting three hypothetical jobs defined by different combinations of attribute levels, from which participants selected their preferred option. The survey also incorporated an internal consistency test and socio-demographic data collection. The two pretests included interviews with volunteer respondents. Results: The first pretest with 67 students (mean age 23.6 ± 2.8 years; 82.1% female) led to wording adjustments for four attributes and two levels. The second pretest with 53 students (mean age 25.3 ± 4.3 years; 67.9% female) confirmed the survey's clarity, lack of hypothetical job dominance, effective trade-off enforcement, and completion time of 15-30 min. Discussion: The second pretest revealed shifts in attribute prioritisation following questionnaire refinements, demonstrating that changes in attribute wording can meaningfully influence perceived importance and highlighting the critical role of precise definitions in shaping respondents' preferences.