|
Dear all,
I am running a hierarchical regression analysis using a correlation matrix
as input using SPSS 11.5, aiming to test the effects of v4 v5 v6 and v7 on
v8 with v1 v2 and v3 controlled. The syntax used is what follows,
MATRIX DATA VARIABLES = ROWTYPE_ v1 v2 v3 v4 v5 v6 v7 v8 .
BEGIN DATA
MEAN 1.698182 3.690909 3.853636 5.434582 4.742327 5.817418 5.566909
4.161212
STDEV 0.459884 0.726775 2.094947 0.578663 0.602034 0.541308 0.516777
0.714999
N 275.000 275.000 275.000 275.000 275.000 275.000 275.000 275.000
CORR 1
CORR -0.08358 1
CORR 0.135497 -0.10633 1
CORR -0.02976 0.004768 0.021313 1
CORR -0.15775 0.052281 -0.02905 0.525542 1
CORR 0.010493 0.051956 -0.00143 0.787962 0.600311 1
CORR -0.00379 0.066051 0.003028 0.720208 0.570858 0.799657 1
CORR -0.05127 -0.05125 -0.00537 0.227996 0.045503 0.233554 0.221487 1
END DATA.
REGRESSION
/MISSING LISTWISE
/STATISTICS COEFF OUTS R ANOVA COLLIN TOL CHANGE
/CRITERIA=PIN(.05) POUT(.10)
/NOORIGIN
/DEPENDENT v8
/METHOD=ENTER v1 v2 v3 /METHOD=stepwise v4 v5 v6 v7.
Unfortunately, I can't get what I want. The results ended up with a one
hierarchy results, not testing the effects of v4-v7.
What's interesting is that the output matrix looks fine as the input one
and when I reperformed a corresponding analysis using part of the syntax,
from "regression" to the end, on the raw data, the results turned out to
be what I want, a clear two hierarchy result.
Could someone give some input on what made the differences?
Thanks a lot in advance!
Wen-Dong Li
Ph. D. candidate
Institute of Psychology, CAS.
|