A Comparison of Formal Methods for Evaluating the Language of Preference in Engineering Design

Abstract

In design, as with many fields, the bases of decisions are generally not formally modeled but only talked or written about. The research problem addressed in this paper revolves around the problem of modeling the direct evaluation of design alternatives and their attributes as they are realized in linguistic communication. The question is what types of linguistic data provide the most reliable linguistic displays of preference and utility. The paper compares two formal methods for assessing a design team’s preferences for alternatives based on the team’s discussion: APPRAISAL and Preferential Probabilities from Transcripts (PPT). Results suggest that the two methods are comparable in their assessment of preferences. This paper also examines the nature of consistency in the way design teams consider the attributes of a design. Findings suggest that assessment of an attribute can change substantially over time.