Re: Getting the Aliasing table for a fractional factorial

[Robert Kinley <KINLEY_ROBERT@LILLY.COM>]

Thanks ... that works a treat

         
      cheers        
       Bob

Moore Terri-ra4974 <Terri.Moore@freescale.com>
Sent by: jmp-l-owner@lists.biostat.wustl.edu

21/07/2005 16:26

Please respond to
jmp-l@lists.biostat.wustl.edu

To	"'jmp-l@lists.biostat.wustl.edu'" <jmp-l@lists.biostat.wustl.edu>
cc
Subject	Re: [jmp-l] Getting the Aliasing table for a fractional factorial

Bob,
you should be able to specify
the full factorial model in jmp then run the Fit Model analysis.  The
'singularity results' section should tell you what is confounded with what
 
Terri
-----Original Message-----
From: jmp-l-owner@lists.biostat.wustl.edu [mailto:jmp-l-owner@lists.biostat.wustl.edu]
On Behalf Of Robert Kinley
Sent: Thursday, July 21, 2005 10:22 AM
To: jmp-l@lists.biostat.wustl.edu
Subject: [jmp-l] Getting the Aliasing table for a fractional factorial
design


Hi 

If I have a datatable for a fractional factorial design 
which was not set up using JMP , is there a way of 
getting its aliasing / confounding table ? 

here's a typical datatable with 3 continuous factors,
2 categorical factors and one response variable. 

temp   time   acid        pos.n    
   spin        response
20        3        44    
   HI        Yes        73
20        3        44    
   LOW        No        521
20        3        220    
   HI        No        55
20        3        220    
   LOW        Yes        77
20        6        44    
   HI        No        39
20        6        44    
   LOW        Yes        58
20        6        220    
   HI        Yes        67
20        6        220    
   LOW        No        941
80        3        44    
   HI        No        23
80        3        44    
   LOW        Yes        1217
80        3        220    
   HI        Yes        193
80        3        220    
   LOW        No        350
80        6        44    
   HI        Yes        201
80        6        44    
   LOW        No        230
80        6        220    
   HI        No        184
80        6        220    
   LOW        Yes        839

I know it's a half-fraction of a 2^5 design, but how 
do I get the actual confounding details ? 

        cheers            
   Bob Kinley