I have used Composite Design (CCD) of Response Surface Methodology (RSM) in one of my set of experiments. When I enter lower and higher factor values, -alpha and + alpha are automatically generated and some of the alpha values are out of range (for example time is negative in numbers). When I go to the CCD table, I could not able to enter some response values due to the impossible factor values of that run. So can I proceed with the missing data for the next steps such as ANOVA? Please help me
Thank you in advance.
Dear Mudila Dhanunjaya Rao,
Literature may be useful to you;
https://online.stat.psu.edu/stat503/book/export/html/681
https://arxiv.org/ftp/arxiv/papers/1603/1603.09716.pdf
https://www.statease.com/docs/v11/tutorials/multifactor-rsm/
Dear Mudila Dhanunjaya Rao,
If you mean that the combinations of the data you have and the experimental design do not match, then it will not be possible to use the RSM. The reason for this is that the experimental design must be created first and then the answers must be processed into the system. Please let me know if I misunderstood your question and you need any further assistance.
Bilge Sayın Börekçi
Dear Bilge Sayın Börekçi,
Yes, I have created an experimental design first and trying to add answers into the system. But those alpha values are not in the range of my factors.
Could you please find the attached Screenshots?
In screenshot 1, the range of lower and higher factors have been entered. But in screenshot 2, I could not able to get some (#std 9 to 14) response data because those factors are not in my range. For example, I cannot perform any experiment with minus 31 minutes time etc. So can I leave those rows and proceed with the analysis?
Please help me
thank you
Dear Mudila Dhanunjaya Rao,
I use the Minitab 17.1.0 software. I also prefer CCD. As far as I can see from the files you uploaded, I processed the factors and factor levels into the system. In these conditions, twenty experiments are applied using a central composite face-centered design (α=1) for the study of three factors each at three levels. When I did it this way, I didn't encounter a problem similar to yours, but I thought maybe this explanation could help you along the way to your study.
Mudila Dhanunjaya Rao
Dear Bilge Sayın Bö,rekçi,
Thank you verymuch for your explanation.
I have chosen a different alpha value by mistake. Therefore I got some strange values. I will proceed with alpha =1.
Thank you
If that's the case, then it might be better to use Box-Behnken Design (BBD) instead of CCD. Or, you may also adjust the range to avoid negative value for minus alpha.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA