Taguchi method with replicates

[Ahernandez2210]

Hi, i got a question, i made an experimental design, and i want to use taguchi method, but i don´t know hot to include the replicates of each experiment in Minitab, somebody know how to do it?

 

[Miner]

One of the drawbacks of using a Taguchi design is that it does not handle replicates. It is intended to analyze repeats. You enter the data for each repeat in a separate column. So, if you have four repeats, create four columns (one for each repeat). You would then calculate an average and a standard deviation of the four repeats for each experimental run. I suppose you could treat your replicates in a similar manner, but Taguchi methods weren't intended for that.

Is there a reason why you don't want to use classical methods? You can achieve the same end result. Simply include the noise factors in the same design.

• Control factors with a linear relationship are scaling factors
• Control factors with a nonlinear relationship are confusion factors
• Control factors with no relationship are economic factors
• Interactions between control factors and noise factors indicate either a robustness factor or a confusion factor depending on the nature of the interaction