Maxdiff and Conjoint

[sur]

Is maxdiff or conjoint is same and what are they and can anybody please help how to write syntax for that.

[David Marso]

Google is a wonderful resource!
Conjoint is part of the Conjoint module/addin.
What is maxdiff?  Probably not the same as Conjoint, but I since you provide no reference....
Why do you want syntax if you don't know what they are?
--

sur wrote

Is maxdiff or conjoint is same and what are they and can anybody please help how to write syntax for that.

[sur]

Hi Sir,

Currently I am a working as a Data Analyst and it is quite difficult to check the  Maxdiff in excel.

So I am wondering if there is a syntax which could help.

Thanks in advance.

[sur]

Can anybody make the logic clear

Thanks for helping me

[sur]

Can anybody helps me in understanding this logic

[Bruce Weaver]

Perhaps someone could if you explained what maxdiff is.  

sur wrote

Can anybody helps me in understanding this logic

[Anthony Babinec]

For a general explanation, see

http://www.sawtoothsoftware.com/products/maxdiff/maxdiff_method


Tony Babinec
[hidden email]

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Bruce Weaver
Sent: Sunday, June 23, 2013 3:56 PM
To: [hidden email]
Subject: Re: Maxdiff and Conjoint

Perhaps someone could if you explained what *maxdiff* is.



sur wrote
> Can anybody helps me in understanding this logic





-----
--

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Rich Ulrich]

That Sawtooth page makes it look rather like one thing that is called
Maxdiff will be a proprietary method.  I see that Wikip has a page, and
what is there says that there can be several different solution methods,
and then names commercial packages that apparently are set up for Maxdiff.
Wikip thus also makes me think "proprietary". 

However, Google Scholar shows that the credited originator, Jordan Louviere,
has numerous popular publications from 1988 or earlier, and a number of
often-cited publications about Maxdiff.  I conclude that you *probably*
can read publications and figure how to do something by that name. 

I guess that I wonder how often people do cobble their own analyses with
other packages.

Someone interested could probably review some of the papers that cite
the technical reports and find out.  If no one happens to already know.

--
Rich Ulrich

> Date: Sun, 23 Jun 2013 22:48:58 -0500

> From: [hidden email]
> Subject: Re: Maxdiff and Conjoint
> To: [hidden email]
>
> For a general explanation, see
>
> http://www.sawtoothsoftware.com/products/maxdiff/maxdiff_method
>
>
> Tony Babinec
> [hidden email]
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
> Bruce Weaver
> Sent: Sunday, June 23, 2013 3:56 PM
> To: [hidden email]
> Subject: Re: Maxdiff and Conjoint
>
> Perhaps someone could if you explained what *maxdiff* is.
>
> ...

[Andy W]

For everyone, I've seen a few posts recently on R-Bloggers that shows the technique is estimatable via an ordinal logistic regression (I am otherwise ignorant though), see http://stats.stackexchange.com/q/48796/1036 and

- http://www.r-bloggers.com/why-doesnt-r-have-a-maxdiff-package/
- http://www.r-bloggers.com/warning-sawtooths-maxdiff-is-nothing-more-than-a-technique-for-rank-ordering-features/
- http://www.r-bloggers.com/if-spss-can-factor-analyze-maxdiff-scores-why-cant-r/

I would implore any analyst to familiarize themselves with a technique enough to explain it (at least alittle!) to strangers before they go and use it.

[Bruce Weaver]

Andy W wrote

I would implore any analyst to familiarize themselves with a technique enough to explain it (at least alittle!) to strangers before they go and use it.

That reminds me of the signature file of a regular in the sci.stat.* groups a few years ago.  I can't now remember who it was, or exactly what the line in the sig file was, but it went something like this:

    "All too often, the analysis of data requires thought."

Maybe Rich or Art will remember who it was.



I am currently out of the office and will return Monday, June 30.  I will have limited access to email and will get back to you as soon as possible. 

Thank you,
McKayla

 

[Bruce Weaver]

I was close, but not quite right.  It was Russell Martin.  Here's his old sig file:

--
All too often the study of data requires care.
The opinions expressed are mine personally and do not
reflect any position of the U.S. Government or NOAA.

Bruce Weaver wrote

That reminds me of the signature file of a regular in the sci.stat.* groups a few years ago.  I can't now remember who it was, or exactly what the line in the sig file was, but it went something like this:

    "All too often, the analysis of data requires thought."

Maybe Rich or Art will remember who it was.

[Eric Black]

Sur

They are not the same.

Conjoint helps you rank attributes and levels (e.g. is color more important
to horsepower and are people willing to pay more for black than blue or red
cars); furthermore you can model market shares given certain scenarios

MaxDiff allows you to rank a large number of attributes on one dimension
only, say which of 35 potential reasons are the most important ones why kids
don't want to clean up their bedroom.

I am not too familiar with the precise estimation, but the kicker for
Sawtooth is that they don't just analyze the result but also help you with
the design.  In MaxDiff you see say 4 reasons and pick the most important
and least important of those four.  You might not want to show certain
attributes on the same screen (if for example one is a subset of the other
or if they are hierarchical).  They then also design the randomized screens
that all the respondents get.

You could probably do all of this manually and within SPSS but I haven't
heard anybody do it.

As far as getting help for writing syntax:  I would assume that this is a
bigger experience than you want to commit yourself to.  Given your original
question I wonder if you are better off with an Off the Shelf product like
Sawtooth.  They have great customer service and will help you out.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD