choice-metrics.com

Hi there,

I am very new to Ngene and am currently designing a DCE. From my understanding I first need to generate a full factorial and although have found the Syntax to do this in the manual I still have a few questions in regards to this:

1. Do I need to upload a data set in order to run the syntax for the full factorial? or do I just run the syntax? Does the dataset need to be in a certain form?
2. I have attempted to run the below syntax but keep getting the error message: Error: the model property contains a prior that has an incorrect number of comma separated elements. 'b'

Design
;alts = alta, altb, altc
;rows = all
;fact
;model:
U(alta) = A[1,2] + B[3, 4, 5, 6] + C[7, 8, 9, 10] + D[11, 12] + E[13, 14] /
U(altb) = A + B + C[15, 16] + D[11, 17] + E[14, 18]
$

1. No, choice experiments in a survey CREATE a data set, no data set is needed.
2. You need to fully specify the utility function, which means it should look like utility = coefficient * attribute + coefficient * attribute + etc. The following syntax works, where I assume that alta and altb are unlabelled alternatives such that all coefficients are generic, and I used C2, D2 and E2 as attribute names because they have different levels from C, D, and E.

Code: Select all

Design
;alts = alta, altb, altc
;rows = all
;fact
;model:
U(alta) = b1*A[1,2] + b2*B[3, 4, 5, 6] + b3*C[7, 8, 9, 10] + b4*D[11, 12] + b5*E[13, 14] /
U(altb) = b1*A + b2*B + b3*C2[15, 16] + b4*D2[11, 17] + b5*E2[14, 18]
$

This results in a design with 8192 rows.

Michiel