topic is it possible if saturated design is efficient too? in SAS Programming

is it possible if saturated design is efficient too?

Khalidjoya — Wed, 31 Jul 2019 12:40:17 GMT

I am doing a research on demand for food safety attributes of vegetables and I tried to design an experiment design through SAS using fractional factorial design (%mktruns) for 5 attributes ( 4 with 2 levels and 1 with 4 levels) so I gained a design with 8 choice sets that is saturated design and efficient too. My question is as we know the saturated design is the smallest design and it is not recommended to use that but in my design the saturated design is efficient and it is recommended by SAS.

I attached a copy of output of the design with this message.

Re: is it possible if saturated design is efficient too?

PaigeMiller — Wed, 31 Jul 2019 13:09:18 GMT

It's not really clear what your question is. The efficiency measures (I think, it's not really clear what efficiency is being calculated) indicate how close to orthogonal the design is.

The design has advantages and disadvantages, which you need to weigh in your mind so you can decide if you want to use it. Efficiency does not negate the disadvantages mentioned.

Advantages:

Small
All main effects estimable
Orthogonal (Efficient)

Disadvantages:

No degrees of freedom for error
No interactions estimable
If interactions are present, then they are confounded with main effects