The overall goal of preference modeling methodologies in the field of designing for human variability (DfHV) is to accommodate (i.e., satisfy the safety and comfort requirements of) the desired percentage of the target user population while concurrently ensuring efficient use of the manufacturer’s resources (time, money, and effort). User anthropometry (body dimensions) play a key role in these accommodation analyses, since they are known to have a significant influence on users’ preferred styles of interacting with products. DfHV’s regression-based approach (RBA) is used to study the influence of user variability (in terms of anthropometry, capabilities, and demographic variables such as age and gender) on the variability of preference within the target population. Results of such studies may be used to make informed design decisions. In contrast, the area of discrete choice analysis (DCA) is replete with methodologies aimed at predicting users’ choices from among given sets of alternatives. Variables used in these choice-prediction models may be alternative-specific (e.g., design specifications, pricing) and/or user-specific (income levels, family size). This research draws on the individual strengths of both the RBA and DCA preference modeling approaches in an effort to develop a more robust design decision-making methodology. The basis of this methodology is the division of the design process into independent analyses of safety and comfort accommodation of the target population. RBA models are recommended for safety accommodation problems while a hybrid of DCA and RBA methods is suggested for the analysis of comfort accommodation. The inclusion of appropriate anthropometric variables in the set of user-specific predictors in DCA models is shown to help enhance the reliability of the models. A case study involving the interaction of a sample population with doorways of a variety of sizes is used to demonstrate and evaluate the application of this decision-making methodology.