I'm performing an experiment with five main effects resp. 6-7 since these main effects influences are studied on 20 sensors
(which can be grouped into 3 manufacturers
).
The design is
Temperature [°C] Rel. Humidity [%] Hydrogen [ppm] Carbon_monoxide [ppm] Ethen[ppm]
50 30 250 250 50
18 11 100 100 100
18 11 500 500 100
18 11 100 500 25
18 11 500 100 25
70 11 100 500 100
70 11 500 500 25
70 11 100 100 25
70 11 500 100 100
50 30 250 250 50
18 84 500 100 100
18 84 100 500 100
18 84 100 100 25
18 84 500 500 25
70 84 500 500 100
70 84 100 100 100
70 84 100 500 25
70 84 500 100 25
50 30 250 250 50
Now, how can I determine which effects are significant and also which interactions and how do they effect the response of the sensors? At the moment I'm trying to model the relationship like given here in How to use multiple variables in a mixed model? Given I find the correct or most meaningful relationship: What's next?