Discrete choice experiments (DCEs) are widely used to elicit consumer preferences, yet how the assignment of attributes and levels affects respondent understanding and preference elicitation remains underexplored. Although prior studies have examined different DCE design factors, little attention has been given to the implications of swapping attributes and levels. This study aims to fill this gap by comparing two designs that reverse the roles of attributes and levels. Our results find that the design using sustainability measures as attributes improves reading ease compared to the design using sustainability outcomes as attributes. Moreover, while both designs show consistent preference patterns across choice sets, variations in willingness-to-pay (WTP) estimates suggest differences in preference consistency between the two designs. These findings highlight that attribute-level assignment may have implications for both respondent and researchers.
