Sensitivity Analysis of Relative Worth in Quality Function Deployment Matrices


This article investigates how sensitive relative worth of product requirements in quality function deployment (QFD) is to different rating scales and/or worth-calculation methods used. The two most popular rating scales and worth-calculation scales are identified from the data of 295 QFD matrices published in conference and journal articles (called empirical QFD), and the sensitivity of relative worth in these empirical QFD (measured by probabilities of large changes of relative worth) is studied by switching rating scales and/or worth-calculation methods. The results suggest that relative worth does not significantly change when a worth-calculation method is fixed. In contrast, if a worth-calculation method is changed, large changes may occur. These large changes are more likely in QFD relationship matrices with small numbers of rows and columns less than or equal to six. These results may imply that designers need to commit to a worth-calculation method when they use information from QFD to make their design decisions. If designers commit to a worth-calculation method, any design decisions made based on relative worth are less likely to be impacted from the use of a rating scale. © 2012 The Author(s).


Mechanical and Aerospace Engineering

Keywords and Phrases

Quality Function Deployment; Rating Scale; Relative Worth; Sensitivity Analysis; Worth Calculation

Document Type

Article - Journal

Document Version


File Type





© 2012 SAGE Publications, All rights reserved.

Publication Date

01 Jan 2012