constant in unlabeled experiments ,and WTP

This forum is for posts that specifically focus on Ngene.

Moderators: Andrew Collins, Michiel Bliemer, johnr

constant in unlabeled experiments ,and WTP

Postby xiaojin » Tue Sep 03, 2019 12:26 am

Dear Ngene experts,
I'm new to Ngene, so I wanted to make sure that I'm doing this right. I'm designing a DCE with 5 attributes (2 with 2 levels, 3 with 2 levels). Each participant will be presented with 6 choice sets. I don't have any information about a priors, so I was trying out zero priors f to do a preliminary experiment and then adjusting the priors based on those initial results. Here is the code I've worked with so far:

Design
;alts = alt1*, alt2*, alt3
;rows = 18
;block=3
;eff = (mnl, d) ;
;model:
U(alt1)=b2.effects[0]*A[0,1]+b3.effects[0]*B[0,1]+b4.effects[0|0]*C[0,1,2]+b5.effects[0|0]*D[0,1,2]+ b6.effects[0|0]*E[0,1,2]/
U(alt2) = b2*A+ b3*B+ b4*C+ b5*D+ b6*E
$

And my question is:
1 I did an unlabeled experiment. Is it correct that the two utility functions in the existing code do not contain constants?
2 The third alternative is a no-choice alternative. Is it necessary to add a constant to it ? Such as “U(alt3) =b1[0]”, What will happen if I don't add it?
3 Attribute E is a cost-related attribute. If WTP is to be calculated eventually, is it necessary to use the code of "; eff = mnl, WTP (wtp1); WTP = wtp1 (b2, b4, b5/b6)"? If I don't add it to the code, will it affect the final analysis of WTP ?

Thank you so much for your help! I am very appreciative.
xiaojin
 
Posts: 20
Joined: Thu Aug 29, 2019 4:55 pm

Re: constant in unlabeled experiments ,and WTP

Postby Michiel Bliemer » Tue Sep 03, 2019 11:52 am

Your syntax looks good.

1 & 2: A no choice alternative is a labelled alternative, therefore you either need to add a constant as you propose in alt3, or you need to add the same constant to both alt1 and alt2. While this will not affect the efficiency of the design much given that your priors are zero, it tells Ngene that you are estimating another parameter, which does matter.

3: You do not need to add wtp in the syntax, you will be able to estimate WTP values anyway. Note that when your cost coefficient is effects coded, you cannot calculate WTPs as simple ratios, you need to calculate WTPs separately for each combination of levels.

Michiel
Michiel Bliemer
 
Posts: 1705
Joined: Tue Mar 31, 2009 4:13 pm

Re: constant in unlabeled experiments ,and WTP

Postby xiaojin » Tue Sep 03, 2019 11:12 pm

Thank you very much. It's very helpful! :D
xiaojin
 
Posts: 20
Joined: Thu Aug 29, 2019 4:55 pm


Return to Choice experiments - Ngene

Who is online

Users browsing this forum: No registered users and 7 guests

cron