[Limdep Nlogit List] NLOGIT command or data construction(No.3)

Mon Aug 13 16:43:22 EST 2007

Hi Hiroshi

I have a quick question, related to your experiment: was the fourth alternative shown to all respondents. That is, is it a no-choice alternative that everyone saw, or did some respondents not see that alternative at all?

If it is a no choice alternative, then your original set-up is okay. You will simply need to not include the attributes in your model utility functions. If some respondents did not see this alternative at all, then the cset should be 3 and the row with alti = 4 should not be present at all within your data. 

John

Dear Prof. Rose, thanks a lot for your guidence.

Here is the result. All data are seemingly read as supposed.
ID: individual respondents
===============================================================================
Variable Mean Std.Dev. Minimum Maximum Cases
===============================================================================
ID 48.5000000 27.7149183 1.00000000 96.0000000 3840
ALTI 2.50000000 1.11817959 1.00000000 4.00000000 3840
CSET 4.00000000 .000000000 4.00000000 4.00000000 3840
CHOICE .250000000 .433069095 .000000000 1.00000000 3840

So as the next step reading page 228-229 in GRH book,
I am wondering if I need to rearrange the dataset.

For example,
My original dataset is this.

  Id, Alti, Cset,choice,place, do tr, price
> 1    1     4    0      4     2   2    2
> 1    2     4    1      1     1   1    6
> 1    3     4    0      2     3   2    2
> 1    4     4    0      0     0   0    0.......none
 the alternative using plece=3 is not in this choice set.

Then do I need to convert like this?
id alti cset choice place do tr price
place1 01 1 5 1 1 1 1 6
place 2 01 2 5 0 2 3 2 2
place3 01 3 5 0 3 ?? ?? ??
place4 01 4 5 0 4 2 2 2
None 01 5 5 0 -999 -999 -999 -999
But if so what value do I need to fill out for "??" for "place3".
Might be -999?

Hiroshi

John Rose wrote;
>
> The cset command informs Nlogit how many alternatives are present within each choice situation. The fact that it is always four simply implies that all choice situations have four alternatives each. It doesn't matter whether the experiment is labelled or unlabled. In the labelled case, all that matters is that for the altij counter 1 is always car, 2 always bus, etc. (substitute whatever labels you have used in practice). In the unlabeled case, alti = 1 is just the first alternative, 2 - the second and so on and so forth.
>
> The error is suggesting that in one or more choice situations, the choice variable has more than one alternative indicated as being chosen. First suggestion is to run the command dstats;rhs= choice$ and see whether the minimum value is 0 and the maximum value 1. It may be that the data has been incorrectly read into the program.
>
>
> John
>
> 
>
> ________________________________
>
> From: limdep-bounces at limdep.itls.usyd.edu.au on behalf of H. KAMEYAMA
> Sent: Mon 13/08/2007 2:14 PM
> To: Limdep and Nlogit Mailing List
> Subject: [Limdep Nlogit List] NLOGIT command or data construction(No.2)
>
>
>
> Dear Prof. Greene,
>
> thank you so much for your reply as usual.
> Now in this case "CSET " is filled out all "4", which means each
> question, choice set, are consisted of four unlabelled choice option.
> Then respondents are required to reply by choosing one of the option in
> each choise set.
>
> As you mention, "my counter, CSET, is incorrect", does it mean due to
> unlabelled case?
> On page 112 in "Applied Choice Analysis" I can see the example, but I
> have no idea how to construct dataset.
>
> The example in the text the first optionis always for "Car", the 2nd for
> "Bas", the 3rd for "Train", the 4th for "Plane",
> whereas in my case the format is different.
>
> Maybe "place" can be the label , but how can I proceed?
>
> Hiroshi KAMEYAMA
>
>
> Prof. Greene wrote,
>  
>> The message means exactly what it says.  For observation
>> (group), 278 the LHS variable does not consist of (0,0,...,1,0,0...).  This
>> usually occurs when your counter, in this case, CSET, is incorrect, so
>> the data set is not correctly blocked for the estimator.
>> /B. Greene
>>
>> 
>> 
>>    
>
>
> --
> ************************************************
> KAMEYAMA, Hiroshi (????)
> Associate Professor
> Farm Management,  Faculty of Agriculture
> Kagawa University, Japan 761-0795
> 2393 Ikenobe Miki-cho, Kita-gun, Kagawa
> ?761-0795 ???????????2393
> ?????????????????????????
> Phone: +81-(0)87-891-3054 (FAX: -3021. admin.,??)
> URL:http://www.ag.kagawa-u.ac.jp/kameyama/index.html
> Email: kameyama at ag.kagawa-u.ac.jp
> ************************************************
>
>
> _______________________________________________
> 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

--
************************************************
KAMEYAMA, Hiroshi (????)
Associate Professor
Farm Management,  Faculty of Agriculture
Kagawa University, Japan 761-0795
2393 Ikenobe Miki-cho, Kita-gun, Kagawa
?761-0795 ???????????2393
?????????????????????????
Phone: +81-(0)87-891-3054 (FAX: -3021. admin.,??)
URL:http://www.ag.kagawa-u.ac.jp/kameyama/index.html
Email: kameyama at ag.kagawa-u.ac.jp
************************************************

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

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: msg-4369-171.txt
Url: http://limdep.itls.usyd.edu.au/pipermail/limdep/attachments/20070813/92ca416d/attachment.txt 