Thanks for the videos! Just wondering, if term A is confounding with interaction term BC, does that mean the main effect we get for A and the interaction effect we get for BC both are useless to us?
The main goal of a resolution III design (the lowest resolution in which main effects are aliased with the two-factor interactions) is to screen important variables. For example, 7 variables can be screened using only 8 experiments. Check the lowest run possible here www.theopeneducator.com/doe/Fractional-Factorial-Design-of-Experiments/Lowest-Runs. Of course, the main effects are indistinguishable from the two-factors interactions. More experimental runs are required to understand that level of details about individual effects (e.g., main, interactions, etc.). However, the systematic fractional design allows using the experiments already run with the new experiments that complement the existing ones, which then allow to distinguish between the main and the interaction effects.
@@TheOpenEducator Thank you for the thought out response. So if two terms are confounding, then we must run more experiments where the factors are not confounding. Got it. Thanks!
Sir can you explain me double confounding and simultaneous confounding
Thanks for the videos! Just wondering, if term A is confounding with interaction term BC, does that mean the main effect we get for A and the interaction effect we get for BC both are useless to us?
The main goal of a resolution III design (the lowest resolution in which main effects are aliased with the two-factor interactions) is to screen important variables. For example, 7 variables can be screened using only 8 experiments. Check the lowest run possible here www.theopeneducator.com/doe/Fractional-Factorial-Design-of-Experiments/Lowest-Runs. Of course, the main effects are indistinguishable from the two-factors interactions. More experimental runs are required to understand that level of details about individual effects (e.g., main, interactions, etc.). However, the systematic fractional design allows using the experiments already run with the new experiments that complement the existing ones, which then allow to distinguish between the main and the interaction effects.
@@TheOpenEducator Thank you for the thought out response. So if two terms are confounding, then we must run more experiments where the factors are not confounding. Got it. Thanks!
2^7 factorial design to create 8 block according which interactions confounded
Find more here: www.theopeneducator.com/doe