Article ID: | iaor20164019 |
Volume: | 32 |
Issue: | 8 |
Start Page Number: | 2977 |
End Page Number: | 2986 |
Publication Date: | Dec 2016 |
Journal: | Quality and Reliability Engineering International |
Authors: | Goh T N, Zhou Min |
Keywords: | design, experiment |
In the traditional use of design of experiments (DOE), the analysis of a designed experiment proceeds only when the designed plan has been fully carried out and the response values are all available, a standard practice that could be inefficient in time and cost and inadequate when quick indicative results are needed. A procedure referred to as ‘iterative designed experiment analysis (IDEA)’ is proposed for suggesting results progressively and giving early indications in situations where response values are available only gradually. The procedure is expected to be useful in the early stages of experimental discovery, especially in factor screening situation. Two case studies illustrate and suggest the potential of the proposed method.