[Limdep Nlogit List] Experimental Design Software for Conjoint

[nmtimet at aragon.es]

Dear Ron,

What do you need exactly? You have to design a choice experiment with various numbers of alternatives and different choice sets where respondents have to choose one alternative form each choice set or you need just to design a conjoint task where respondents have to rank or order choices from a unique choice set?

Nadhem

----- Mensaje original -----
De: infocorp <infocorp at ozemail.com.au>
Fecha: Miércoles, Octubre 25, 2006 12:47 pm
Asunto: [Limdep Nlogit List] Experimental Design Software for Conjoint

> Dear Limdep Users,
> 
> Is anyone able to recommend affordable experimental design generation
> software which directly generates conjoint designs when attributes 
> havevarying numbers of levels.
> 
> An hypothetical example of a conjoint task is:
> 
> 1 x 2 level attribute
> 1 x 3 level attribute
> 1 x 4 level attribute
> 1 x 5 level attribute
> 1 x 6 level attribute
> 1 x 7 level attribute
> 
> SAS QC suggests generating designs with PROC FACTEX using powers of
> prime numbers and then collapsing levels.
> 
> The 2 level attribute would be generated directly; the 3 level 
> attributewould be generated by interacting two 2 level attributes 
> to give 4
> levels, 2 of which would be collapsed; and so on until the 7 level
> attribute would be generated by interacting three 2 level 
> attributes to
> give 8 levels, 2 of which would be collapsed.
> 
> Is there software available which will produce a design directly from
> the specification above?
> 
> Ron
> 
> 
> 
> 
> 
> _______________________________________________
> Limdep site list
> Limdep at limdep.itls.usyd.edu.au
> http://limdep.itls.usyd.edu.au
>