Durbin's test for Incomplete Block Design

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Durbin's test for Incomplete Block Design

Dussault, Andre [JNJCA]
Hi, I'd like to perform the Durbin's test when using Incomplete Block
Design. It is a modification of Friedman's test.

I don't know how to perform this type of test in SPSS, does anyone knows a
macro or a syntax to do so. Note that this is not the classic Durbin-Watson
sign test.

Thanks in advance

-----Message d'origine-----
De : SPSSX(r) Discussion [mailto:[hidden email]]De la part de
Richard Ristow
Envoyé : Monday, December 11, 2006 1:22 PM
À : [hidden email]
Objet : Re: Ranking within cases individually


A correction - Marta, I'm glad I read your postings.

At 02:44 AM 12/11/2006, Marta García-Granero
wrote ("Re: General Jackknife Syntax"):

>Trying to run the code with datasets more than
>40 cases long will trigger a MATRIX error
>concerning MAX LOOPS being reached. It can be
>avoided using SET MXLOOPS=100 (or 1000...
>depending on sample size) before the MATRIX...END MATRIX procedure.

This applies to the number of SPSS 'cases' - rows
in file, plots in your study - as well. I recall you have about 12,000.