Help with MaxDiff

[Auberth Hurtado]

Hello, somebody have information (for beginners) of Maximum Difference
Scaling (MaxDiff) analysis, I looked on internet and Sawtooth software
pages, but I still stuck.

I understand that analysis can be done in SPSS but I can't understand how.

Thanks.



Auberth Hurtado.

=====================
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

[Abdus Salam]

Hi ,

  As per I know Maximum difference scaling analysis which called also best -worst analysis
  its only possible in swatooth software MX-difference software. and to check the impotence
  i.e. to extract the standardized beta-coefficient you need also CBC-HB ;so that you can calculate the importance.

  you can make the orthogonal design in SPSS though its better to use Mx-difference design.

  Thanks!
  Mousumi.






Auberth Hurtado <[hidden email]> wrote:

  Hello, somebody have information (for beginners) of Maximum Difference
Scaling (MaxDiff) analysis, I looked on internet and Sawtooth software
pages, but I still stuck.

I understand that analysis can be done in SPSS but I can't understand how.

Thanks.



Auberth Hurtado.

=====================
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



---------------------------------
Be a better sports nut! Let your teams follow you with Yahoo Mobile. Try it now.

=====================
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

[Brandon Paris]

Hi Auberth --

MaxDiff can be designed as an incomplete block design based on the
number of items you wish to include.  Unfortunately, I am not familiar
on how to do this in SPSS, as there is only 1 variable in the study.
But, there are statistics texts that provide solid incomplete block
designs -- a good one is Experimental Designs by Cochran and Cox.

Once you have the design and have collected the data, you can:

1) For each set and each person, assign the item a +1 if it the item was
chosen as best, -1 if the item was chosen as worst.  Then, for each
item, add up the total (all +1's and all -1s).  The sums represent the
score on the new scale.  This is what Louviere originally described, I
believe.  As long as your design is balanced, you can do this in
aggregate and for each respondent.

2) Run an aggregate multinomial logit to estimate the utilities (you
will get only one utility per item in the study using this method -- no
respondent-level utilities/scores).  This takes some manipulation of the
design file, however.

3) Estimate using hierarchical Bayes techniques, but that requires
another stat package, such as Sawtooth, R, or S+.  To my knowledge, you
can't do this in SPSS.  Like (1), you'll also get results in aggregate
and for each respondent.


Thanks,
Brandon
[hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Auberth Hurtado
Sent: Tuesday, November 20, 2007 4:13 PM
To: [hidden email]
Subject: Help with MaxDiff

Hello, somebody have information (for beginners) of Maximum Difference
Scaling (MaxDiff) analysis, I looked on internet and Sawtooth software
pages, but I still stuck.

I understand that analysis can be done in SPSS but I can't understand
how.

Thanks.



Auberth Hurtado.

=====================
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

=====================
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