Taguchi’s approach to Design of Experiments (DOE) is a very broad subject, and we won’t be able to cover everything. We will focus on Taguchi DOE as a special type of fractional 2^k design of experiments (although we will also consider 3^k, 4^k, and 5^k designs).
One of the goals of the Taguchi approach is to optimize the early stages of product or system design, especially by reducing the number of experiments or trials that need to be conducted. Here the goal is to extract as much information as possible for the least cost (in terms of time, resources, and money). The approach uses a small number of standard orthogonal arrays and fits the design to one of these arrays.
Topics
- 2-level design basics
- Design optimization
- List of 2-level designs and interactions
- Real Statistics support
- Design with replications
- Signal-to-noise ratio
- 3-level designs
- 3-level examples
- 4-level designs
- Interactions for 4-level designs
- 5-level designs
References
Roy, R. K. (2010) A primer on Taguchi method. 2nd ed. Society of Manufacturing Engineers
https://www.scribd.com/document/381615601/288020391-a-Primer-on-the-Taguchi-Method
University of York (2004) Orthogonal arrays (Taguchi designs)
No longer available online
Minitab (2024) Catalogue of Taguchi designs
https://support.minitab.com/en-us/minitab/help-and-how-to/statistical-modeling/doe/supporting-topics/taguchi-designs/catalogue-of-taguchi-designs/
Kacker, R. N., Lagergren, E. S. and Filliben, J. J. (1991) Taguch’s orthogonal arrays are classical designs of experiments. J. Res. Natl. Inst. Stand. Technol. 96, 577
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4927234/pdf/jresv96n5p577_a1b.pdf
Hi, I am currently using the Taguchi DOE feature from the ANOVA tab. I selected an L16 (4) orthogonal array in Excel, but every time I attempt to run the analysis, I receive the following error: “No Taguchi design is specified.” I have selected 5 factors along with their headers and have tried both my actual level values and the coded values (1–4) according to the standard pattern. I am using the macOS version of the software.
Could you please advise on how to resolve this issue so that the Taguchi design can be properly recognized? Thank you in advance for your assistance.
Hello Victoria,
Sorry for the delayed response. I will try to test this on a Mac in the next day or two.
Charles
Hi Victoria,
1. I checked and this Taguchi DOE model does run correctly on my Mac.
2. I understand that you are using the L16 (4) design with no interactions. Suppose we use the deisgn in Figure 2 of
https://real-statistics.com/design-of-experiments/taguchi-design-of-experiments/4-level-taguchi-designs/
Thus you must insert something like the letters A B C D E in the row next to Fact, i.e range B2:F2.
You also need to place numeric values in the column to the right of the design array, namely range G3:G18
Is this what you did or are you trying to do something different?
Charles