OOD designs, level specification and D-optimality
Posted: Thu Dec 02, 2010 4:28 amHi there,
I'm trying to run an OOD design for a small choice experiment, consisting of 2 profiles, each consisting of two 3-level attributes and one 9-level attribute.
My code is below:
The design appears to be generated fine, but the D-Optimality remains "Undefined". I find this to be the case unless I keep all the attributes below 5 levels. Is this a known limitation of the software?
I'm trying to run an OOD design for a small choice experiment, consisting of 2 profiles, each consisting of two 3-level attributes and one 9-level attribute.
My code is below:
Design
;alts = alt1, alt2
;rows = 18
;block = 3
;orth = ood
;model:
U(alt1) = b0 + b1.dummy[0|0] * A[0,1,2] + b2.dummy[0|0] * B[0,1,2] + b3.dummy[0|0|0|0|0|0|0|0] * C[0,1,2,3,4,5,6,7,8] /
U(alt2) = b1 * A + b2 * B + b3 * C $
The design appears to be generated fine, but the D-Optimality remains "Undefined". I find this to be the case unless I keep all the attributes below 5 levels. Is this a known limitation of the software?