Sensitivity Analysis in Perceptual Evaluation of System Module Concepts

Selecting optimum concepts for a system and its subsystems in the conceptual design stage involves uncertainties due to imperfect information about customer preferences (market shares), cost of the system developed from each concept, and feasibility of new technology used in the new system. When analytical relationships between system performance and system inputs or parameters are unknown in the early system development stage, one approach to quantify the goodness of a concept is to use rating scales. This paper studies the effects of variations (precisions) in rating scales and in cost estimation for evaluating the goodness of system module concepts (e.g., sub-systems, assemblies, subassemblies, and parts). This paper presents a global sensitivity analysis (GSA) in perceptual concept evaluation, three probability measures for evaluating and selecting optimum concepts in GSA, and one-set-of-factors-at-a-time GSA to identify the sets of factors that cause significant variations in concept evaluation outcomes.