DOE approach

[#1]

Hi,
Recently i have been asked if there is a reference for the following statement :
DOE approach is the optimal (best/only effective?) way to conduct experiments that will lead to data that can be analyzed and to a model or correlation and introductions.

Would appreciate it if someone can share an academic reference for thia kind of statement.

Thanks

[#2 · Accepted solution]

Hi @YanivD,

No doubt about this statement !
DoE is the best (and most efficient) method to get informations with a minimum number of experiments, and in some cases the only way to get specific informations like interactions between factors. It's a methodology that can also support sequential experimentation (through Design augmentation), so it's highly adaptable depending on your initial findings/results.
You can find different publications aiming to demonstrate the use and effectiveness of DoE. Among (many !) others, here are some publications or links I found:
- An example with a comparison between OFAT and DoE approach : https://www.benchmarksixsigma.com/forum/topic/35533-one-factor-at-a-time-ofat-experimentation/
- "A Design of Experiments (DoE) Approach Accelerates the Optimization of Copper-Mediated 18F-Fluorination Reactions of Arylstannanes" (take a look at Figure 2) : https://doi.org/10.1038/s41598-019-47846-6
- "Improvement of OFAT through DoE" 10e.xps (psu.edu)
- OFAT Vs DoE : D:DVICZITROM.DVI (psu.edu)

There are a lot of publications demonstrating the relevance and effectiveness of DoE, you can try searching for these papers on Google (or other search engines) by searching "comparison OFAT vs. DoE" or similar searches. There are also a lot of examples in the Community about successful DoEs, so take a look at JMP Discovery papers and presentations.

Hope this first answer will help you,

[#3]

Thanks a lot! Great

[#4]

I also recommend the book, "Optimal Design of Experiments: A Case Study Approach," by Peter Goos and Bradley Jones, if you are going to use JMP to design and analyze your experiments.

[#5]

More papers for folks new to DOE approach:

https://williamghunter.net/george-box-articles/statistics-for-discovery

https://williamghunter.net/george-box-articles/statistics_as_a_catalyst_to_learning

https://williamghunter.net/george-box-articles/the-scientific-context-of-quality-improvement

I'm obviously a fan of G.E.P. Box...

I will add, there are other ways to get data to provide insight to causal relationships.  When you have a lot of factors to consider, directed sampling (component of variation) studies are quite useful.

[#6]

Use Google Scholar and Google DOE, or more specfically JMP, DOE. Its getting diffcult to keep up with the number of papers that use DOE in my field. In general the use of DOE in academic publications has grown alot.

About 4,660 results in 2022