Sensitivity of Relative Worth in QFD Matrices to Rating Scales and Worth Calculation Schemes: An Empirical Study

In the past several decades, Quality Function Deployment (QFD) has gained its popularity among engineers as a tool to relate (map) customer requirements (inputs) to system requirements or system components (outputs), and to calculate relative worth of these requirements and components. The benefits of QFD in system development include cost reduction, fewer design changes after the start of production, and improved communication among engineers. Despite the observed benefits, the needs for QFD research have been addressed by researchers. These research needs include decision-making process in QFD, rating scales used in QFD matrices, and calculation schemes for calculating the worth of outputs (e.g., system requirements) from the importance of inputs (e.g., customer requirements). The purpose of this paper is to empirically study sensitivity of the relative worth and rank of outputs in QFD matrices to rating scales and worth calculation schemes. We collected QFD matrices from journal articles and textbooks; then calculated the changes of relative worth and the rank of outputs when one type of rating scale and/or worth calculation scheme was changed to the other. The results suggest that the relative worth and rank of outputs are relatively insensitive to rating scales and worth calculation schemes in the QFD matrices studied in this paper