The constant in the syntax.

This forum is for posts that specifically focus on Ngene.

Moderators: Andrew Collins, Michiel Bliemer, johnr

The constant in the syntax.

Postby sukunta » Fri Feb 14, 2020 2:18 am

Dear Prof.Michiel Bliemer,

My generic DCE with neither choice of model for the pilot study has two attributes with two dummy levels, one attribute with three dummy levels and two attributes with three continuous levels. My syntax as follows
Code: Select all
design
;alts=alt1*,alt2*,alt3
;rows=12
;eff=(mnl,d)
;con
;model:
U(alt1)=b0[0.000001]+b1.dummy[0.000001]*format[1,0]+b2.dummy[0.000001]*con_ser[1,0]+b3[-0.000001]*time[10,20,30]+b4.dummy[0.000001|0.000002]*area[1,2,0]+b5[-.000001]*cost[0,50,100]/
U(alt2)=b1*format+b2*con_ser+b3*time+b4*area+b5*cost
$


and my questions are
1. In the manual page 100 "Ngene ignores any constants" I do not know about
this point. please clarify it to me. I include the constant in my syntax that is correct or not?
2. The K includes constant then the number of the row should be 7 rows and
to supporting level balance, thus 12 rows are suitable. If the syntax should not include the constant then the rows are 6. My understanding is correct or not?
Sincerely yours,
Sukunta Muadthong
sukunta
 
Posts: 80
Joined: Tue Jan 12, 2016 1:28 pm

Re: The constant in the syntax.

Postby Michiel Bliemer » Fri Feb 14, 2020 8:17 am

1. The manual states: "in calculating the efficiency measures, Ngene ignores any constants". This does not mean that Ngene ignores the constants in the utility function, but rather it does not optimise for estimating the constants when minimising the D-error. If you want to also optimise for constants, you include the command ;con. Note that you indicate that alt1 and alt2 are generic, via alt1*,alt2*, but you do include a constant. Unlabelled alternatives (which have generic utility functions) cannot be different, so you need to use the same utility function for both alternatives. You can add a constant in model estimation, but then this constant only refers to left-to-right bias and should not be interpreted with respect to preferences.

The appropriate syntax would be:

Code: Select all
design
;alts=alt1*,alt2*,alt3
;rows=12
;eff=(mnl,d)
;con
;model:
U(alt1)=b0[0.000001]+b1.dummy[0.000001]*format[1,0]+b2.dummy[0.000001]*con_ser[1,0]+b3[-0.000001]*time[10,20,30]+b4.dummy[0.000001|0.000002]*area[1,2,0]+b5[-.000001]*cost[0,50,100]/
U(alt2)=b0[0.000001]+b1*format+b2*con_ser+b3*time+b4*area+b5*cost
$


In other words, I added the same constant to alt2, while U(alt3)=0, so the constant now refers to selecting an alternative against making no choice.

2. S*(J-1) >=K, where K = 7 in your case. Since J = 3, this means that S >= 4. You can try this in Ngene, if you set ;rows = 3 then Ngene will tell you that you need to use 4 rows in order to satisfy the degrees of freedom.

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

Re: The constant in the syntax.

Postby sukunta » Fri Feb 14, 2020 12:09 pm

Dear Prof. Michiel Bliemer,

Thank you so much. Your explanation is very clear to me.

Sincerely yours,

Sukunta Muadthong
sukunta
 
Posts: 80
Joined: Tue Jan 12, 2016 1:28 pm


Return to Choice experiments - Ngene

Who is online

Users browsing this forum: Bing [Bot], Google [Bot] and 6 guests

cron