Well, a contrast is a weighted function of the means set equal to zero
such that the sum of the contrast weights is also zero. A comparison is
usually reserved for pairwise comparisons. Thus a comparison is a
contrast but a contrast may be more than a comparison, a complex
contrast. The planned term means it is hypothesized before the fact. The
most efficient method is planned orthogonal contrasts. They are a priori
and do not ask redundant questions. Using all possible pairwise
comparisons is redundant. Pairwise comaprisons are usuallly done
following a significant omnibus test. planned contrasts are done instead
of an omnibus test.
Paul R. Swank, Ph.D.
Professor, Developmental Pediatrics
Director of Research,
University of Texas Health Science Center at Houston
-----Original Message-----
From: SPSSX(r) Discussion [mailto:
[hidden email]] On Behalf Of
KEVIN MANNING
Sent: Thursday, October 26, 2006 3:13 PM
To:
[hidden email]
Subject: Planned Comparison vs Contrast
Dear list,
My first time. Forgive my ignorance. I have noticed in the
literature authors discuss using planned comparisons and planned
contrasts with repeated measures ANOVA.
I cannot seem to find a disucssion of the difference or when to use
either of them -
I am using planned comparisons in the syntax below but am not 100% sure
this is appropriate
Any info or where to look would be deeply appreciated. Thank you,
Kevin
GLM
var 1 var2 BY group
/WSFACTOR = test 2
/CONTRAST (group)=Simple
/METHOD = SSTYPE(3)
/EMMEANS = TABLES(group) COMPARE ADJ(LSD)
/EMMEANS = TABLES(test) COMPARE ADJ(LSD)
/EMMEANS = TABLES(group*test)
/EMMEANS = TABLES(group*test) compare (test) adj(bonferroni)
/EMMEANS = TABLES(group*test) compare (group) adj(bonferroni)
/PRINT = DESCRIPTIVE ETASQ TEST(MMATRIX)
/CRITERIA = ALPHA(.05)
/WSDESIGN = test
/DESIGN = group.
Locked
