choice-metrics.com

Dear all THE FOLLOWING did not include the intercept (constant) therefore, is the following acceptable to run the efficient design ..tq

Design
;alts = alt1*, alt2*
;rows = 12
;eff = (mnl,d)
;block = 2
;model:
U(alt1) = b1[0.17]*A[0, 1, 2] + b2[0.18]*B[0,1,2] + b3[0.13]*C[0,1,2] + b4[0.13]*D[0,1,2] + b5[-0.2]*E[0,1,2,3] /
U(alt2) = b1*A + b2*B + b3*C + b4*D + b5*E$

Yes this is definitely acceptable. If you have unlabelled alternatives, which you seem to have, then you typically do not include an alternative specific constant.

Thank you for your kind reply ( Michiel Bliemer ).