topic Re: How to assess factor effects in a Mixture-in-mixture experiment? in Discussions

How to assess factor effects in a Mixture-in-mixture experiment?

frankderuyck — Tue, 24 Feb 2026 09:36:10 GMT

I have three formulations A, B and C each is a mixture of three components c1, c2 and c3; the component fraction in each formulation is different. How to design a three formulation A/B/C mixture experiment so that the effect of each component c on a response variable Y can be assessed?

Re: How to assess factor effects in a Mixture-in-mixture experiment?

Victor_G — Wed, 25 Feb 2026 09:35:05 GMT

Hi @frankderuyck,

I'm not sure about your question ?

If you're interested about designing a mixture of mixture scenario, you can check the example here:
If you're interested about assessing relative factors importance on a response variable, you can check the parameters estimates from the fitted model, or assess variable importance through prediction profiler using Dependent Resampled Inputs or Linearly Constrained Inputs method (both methods should provide similar factor importance values).

Hope this answer will help you,

Re: How to assess factor effects in a Mixture-in-mixture experiment?

frankderuyck — Wed, 25 Feb 2026 10:44:14 GMT

Hi Victor, nice example!

Think there is a mistake in the second onstraint: Butter + Milk + Eggs </= 0,55?

How to design when also the wet & dry ingredients also are variable mixture components e.g %Dry 30 - 45 and %Wet 40 - 55 ?

Re: How to assess factor effects in a Mixture-in-mixture experiment?

frankderuyck — Wed, 25 Feb 2026 10:47:25 GMT

Correction wet a dry limits above: what if %Dry 30 - 45 and %Wet 55 - 70 ?

Re: How to assess factor effects in a Mixture-in-mixture experiment?

Victor_G — Wed, 25 Feb 2026 11:11:50 GMT

In this example, dry components should sum up to exactly 45%, hence the double constraints with 0,45.

But if you want to vary the dry content between 30 and 45%, you simply have to change the lower limit value to 0,30 in the second inequality constraint.