SAS Note V7-CATMOD-0347

From: tsdlist@UNX.SAS.COM
Date: Thu Feb 11 1999 - 15:53:45 EST


>>>>>>>> SAS Notes -- Copyright (c) 1991 - 1999 by SAS Institute Inc. <<<<<<<<

V7-CATMOD-0347 *** ALERT NOTE *** PRODUCT: STAT
                                                        PROCEDURE: CATMOD
TITLE: Incorrect tests of between*within effects in repeated measures models

KEYS: INCORRSTAT REPEATED INTERACTIONS BETWEEN BY WITHIN ORDER ESTIMATES
       CHANGE TESTS

    If you specify more than one within-subjects factor in the REPEATED
    statement (e.g. REPEATED TIME 2, TRIAL 2;), and if you specify
    interaction(s) of between- and within-subjects factors on the MODEL
    statement (e.g. group*_response_), then the tests of these interactions
    will be incorrect in the Analysis of Variance table. Also, the
    parameter estimates, while correct, are not correctly organized in the
    Analysis of Weighted-Least-Squares Estimates table. PROC CATMOD
    generates the correct design matrix columns, but if a between*within
    interaction requires more than one column, those columns are not
    consecutive in the matrix. Consequently, they are not in the order
    stated in the Estimates table and the wrong contrast of parameters is
    tested in the Analysis of Variance table. By examining the design
    matrix, you can find the columns belonging to the interaction and then
    produce a correct test of it using the CONTRAST statement. One symptom
    of this problem is that tests of these interactions change if you change
    the order of the within-subjects factors in the _RESPONSE_= option of
    the REPEATED statement.

+------REPORTED------+ +---FIXED---+
SYSTEM RELEASE LEVEL RELEASE LEVEL

AIX/R 7.00 0P1 .
CMS 7.00 0P1 .
DigitUnx 7.00 0P1 .
HP800 7.00 0P1 .
MVS 7.00 0P1 .
OS/2 7.00 0P1 .
VMS 7.00 0P1 .



This archive was generated by hypermail 2b29 : Mon Jan 10 2000 - 10:10:19 EST