[Limdep Nlogit List] Discrete choice

[Mujahed Mohammad Thneibat]

Thanks Behram,

I'm interested to know "which devices to buy". Basically, I want to see the attributes of different devices on the utility if attaining them. 

What do you think of a probit model for this situation?

Kind regards,

Mujahed

----- Original Message -----
From: "Behram Wali" <walibehram at yahoo.com>
To: "Limdep and Nlogit Mailing List" <limdep at limdep.itls.usyd.edu.au>
Sent: Sunday, January 17, 2016 7:51:47 PM
Subject: Re: [Limdep Nlogit List] Discrete choice

"Which devices to buy" or "how many to buy"?If its the later one, then there exists intrinsic ordering pattern within your response variable, 
and modeling such a response variable via unordered framework is likely to result in loss of important information. Ordered probit/logit in some cases may not give best fit for all ordered response variables. However, on a general note, ordered (probit and/or logit) model would be an intuitive methodological alternative in your case.

    On Tuesday, January 12, 2016 10:39 PM, Mujahed Mohammad Thneibat <mthneiba at purdue.edu> wrote:
 

 Hi there, 

I have a discrete choice model where respondents are asked to specify which devices they would buy.

Most of the respondents have indicated that they will buy more than two devices (out of six devices). 

Is there a way to make a discrete choice model? or Should I go for a probit model? 

Any feedback is welcomed. Thanks in advance for the help.

Mujahed 

Lyles School of Civil Engineering
Purdue University 
_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au


   
_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au