What Statistical Designs should I use?

[bbh]

Hi! We're making a Design of Experiment that involves the use of rubber band and egg shell to see of it affect compression strength of a concrete.

So, we make a proportion of the cement and two control variables in interval like
1. 100:0:0
2. 50: 50 :0
3. 50: 0: 50
4. 50: 25: 25
5. 50: 75: 25
6. 50: 25: 75


the cement is held constant. What design should I use to determine the best proportion that will yield an optimum result? Thank you

 

[Miner]

It depends on your goal. If you are only interested in which of the 6 formulations is best, you can run those and analyze using a 1-way ANOVA. If you want to determine which formulation (not necessarily one of the 6) is best as well as how each ingredient contributes and possibly interacts with other ingredients, I recommend a mixture design.

 

[bbh]

thanks for the help!