Can SPSS fit a Bradley-Terry model

[drfg2008]

yes it can.

However, I found an example on IBM / SPSS homepage with syntax but without data. Does anyone know where the data could be. Have I missed something?

https://www-304.ibm.com/support/docview.wss?uid=swg21476112


Frank

[Bruce Weaver]

You may be able to get the data via the website for Agresti's book.

   http://www.stat.ufl.edu/~aa/cda/cda.html

drfg2008 wrote

yes it can.

However, I found an example on IBM / SPSS homepage with syntax but without data. Does anyone know where the data could be. Have I missed something?

https://www-304.ibm.com/support/docview.wss?uid=swg21476112


Frank

[drfg2008]

thanks, ... as a SAS data file



data baseball;
input wins games milw detr toro newy bost clev balt;
datalines;
 7 13  1 -1  0  0  0  0  0
 9 13  1  0 -1  0  0  0  0
 7 13  1  0  0 -1  0  0  0
 7 13  1  0  0  0 -1  0  0
 9 13  1  0  0  0  0 -1  0
11 13  1  0  0  0  0  0 -1
 7 13  0  1 -1  0  0  0  0
 5 13  0  1  0 -1  0  0  0
11 13  0  1  0  0 -1  0  0
 9 13  0  1  0  0  0 -1  0
 9 13  0  1  0  0  0  0 -1
 7 13  0  0  1 -1  0  0  0
 7 13  0  0  1  0 -1  0  0
 8 13  0  0  1  0  0 -1  0
12 13  0  0  1  0  0  0 -1
 6 13  0  0  0  1 -1  0  0
 7 13  0  0  0  1  0 -1  0
10 13  0  0  0  1  0  0 -1
 7 13  0  0  0  0  1 -1  0
12 13  0  0  0  0  1  0 -1
 6 13  0  0  0  0  0  1 -1
;
proc genmod;
   model wins/games = milw detr toro newy bost clev balt /
   dist=bin link=logit noint covb;

[Rolf Pfister-2]

Guten Tag

ich bin bis am 8. August 2011 nicht im Büro und lese die Emails erst, wenn ich zurück bin. In dringenden Fällen wenden Sie sich bitte an [hidden email] oder an den Support ([hidden email]).

Vielen Dank für Ihr Verständnis

 

Hello,

I'm not in the office until 8 August 2011 and I will read my emails only when I'm back. For urgent cases, please contact [hidden email] or our support ([hidden email]).

Thank you for your comprehension.

 

[drfg2008]

that was a bit premature,

it's not the data

[vlad simion]

Hi Frank,

that is the data from Agresti's example of paired comparison model, but is in the aggregated form. 

For the example from IBM/SPSS site you need to tweak a little the data file and restructure it in the long form and then code for the unique pairs of teams as explained in that post. 

Here is a small syntax that gives you the table from Agresti's example:

data list free 

/ winner loser cnt cwt. 

begin data 

1,2,7,1,1,3,9,1,1,4,7,1,1,5,7,1,1,6,9,1,1,7,11,1,2,3,7,1,2,4,5,1,2,5,11,1,2,6,9,1,2,7,9,1,3,4,7,1,3,5,7,1,3,6,8,1,3,7,12,1,4,5,6,1,4,6,7,1,4,7,10,1,5,6,7,15,7,12,1,6,7,6,1,2,1,6,1,3,1,4,1,3,2,6,1,4,1,6,1,4,2,8,1,4,3,6,1,5,1,6,1,5,2,2,1,5,3,6,1,5,4,7,1,6,1,4,1,6,2,4,1,6,3,5,1,6,4,6,1,6,5,6,1,7,1,2,1,7,2,4,17,3,1,1,7,4,3,1,7,5,1,1,7,6,7,1,1,1,0,0,2,2,0,0,3,3,0,0,4,4,0,0,5,5,0,0,6,6,0,0,7,7,0,0

end data.

format all (f4.0).

weight by cnt.

ctables

  /table winner[c] by loser[c].

Hth,

Vlad

On Fri, Jul 22, 2011 at 9:10 PM, drfg2008 <[hidden email]> wrote:

that was a bit premature,

it's not the data

-----
Dr. Frank Gaeth
FU-Berlin

--

View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Can-SPSS-fit-a-Bradley-Terry-model-tp4623573p4624126.html

Sent from the SPSSX Discussion mailing list archive at Nabble.com.

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

[drfg2008]

thanks vlad for the data

[drfg2008]

great

thanks for the new data + syntax, vlad.

[Nicolas]

Could someone show me 2 or 3 lines as an example of how should be exactly the table in SPSS. I mcannot understand correctly the whole procedure.
Thanks in advance