Skip to main content
. Author manuscript; available in PMC: 2010 Dec 6.
Published in final edited form as: J Nurs Meas. 2009;17(2):91–104. doi: 10.1891/1061-3749.17.2.91

Table 3.

Goodness-of-Fit Tests for Alternative Measurement Models

χ 2 df χ2/df Δ χ 2 GFI CFI BBNFI RMSEA
Two-Factor Models (19 CES-D items)
without cross-group (black vs. white) constraints:
1) base model 1
 (1 factor loading for each item, covariance between Factor 1 & 2)
666.44 302 2.21 --- 0.899 0.970 0.966 0.40
2) base model 2
 (model 1 plus covariance between CES-D 15 & 19)
598.25 300 1.99 68.19* 0.909 0.975 0.972 0.36
3) base model 3
 (model 2 plus additional factor loading of CES-D 12 on Factor 1)
575.27 298 1.93 22.55* 0.912 0.977 0.974 0.35
4) base model 4
 (model 3 plus additional factor loading of CES-D 6 on Factor 2)
559.95 296 1.89 15.32* 0.914 0.978 0.975 0.35
5) base model 5
 (model 4 plus additional factor loading of CES-D 8 on Factor 1)
541.57 294 1.84 18.38* 0.916 0.980 0.976 0.34
Two-Factor Models (19 CES-D items)
with cross-group (black vs. white)constraints:
1) constrained model 6
 (model 5 plus cross-group constraints on all factor loadings and covariances)
569.41 316 1.80 27.84$ 0.913 0.979 0.977 0.33
2) constrained model 7
 (model 6 plus release of cross-group constraint on factor loading of item CES-D 12)
561.91 315 1.78 7.50* 0.914 0.980 0.978 0.32
*

p<0.05

$

p≥0.05