Page 1 of 1

### which is the prior for b0 in U(alt3)

Posted: Fri Dec 06, 2019 12:52 pm
Dear Ngene Experts,
Here is our code for DCE pilot design. After piloting, we analyzed the pilot data and used asclogit in Stata to analyze the pilot data. The stata result is as follows. The question is for the formal design, which is the prior from Stata asclogit result should be used for b0 in U(alt3)?

Design
;alts=alt1*,alt2*,alt3
;rows=24
;block=4
;eff=(mnl,d)
;model:
U(alt1)=b1.effects[0.001]*A[1,0]+b2[0.001]*B[3,2,1,0]+b3.effects[0.001|0.001|0.001]*C[3,2,1,0]+b4[0.001]*D[0,20,40,60]+b5[-0.001]*E[100,250,400,550]/
U(alt2)=b1*A+b2*B+b3*C+b4*D+b5*E/
U(alt3)=b0[0]
\$

The stata result(coef):
Alt : b1:.3790657
b2:.0002358
b3 :.0411187; 0.1913767; 0.1317095
b4 :.0083367
b5 :.0017246

1 cons: 2.490288
2 cons: 2.40686
3: (base alternative)

Thank you very much!

### Re: which is the prior for b0 in U(alt3)

Posted: Mon Dec 09, 2019 10:22 am
I do not really understand the output of Stata. Do I assume correctly that "cons" are referring to the constants for al1 and alt2? Then it would essentially be the following (where I am using a generic constant for alt1 and alt2 equal to the average):

U(alt1) = b0[2.45] + b1...
U(alt2) = b0 + b1...
U(alt3) = 0

In other words, this can be rewritten as:

U(alt1) = b1...
U(alt2) = b1...
U(alt3) = b0[-2.45]

In this case you would choose -2.45 as the prior for b0.

Michiel