Skip to main content
. 2010 Oct 25;8:120. doi: 10.1186/1477-7525-8-120

Table 3.

Results of tests of invariance for the VETERAN (N = 314) and DOPPS (N = 3,300) samples

Comparative Statistics

Model χ2 df Contrast with Model # Δχ2 Δdf Unadj. p < Bonf. Adj. p < ΔCFI w2
1. Baseline model: Two factors (KDCS & SF-36) with no invariance constraints 2796.225 178 - - - - - - - - - - - - - -

2. KDCS factor loadings invariant 2819.092 184 1 22.867 6 .00085 .025 .0003 .08

3. SF-36 factor loadings invariant 2804.771 185 1 8.546 7 .29 ns .0002 .05

4. KDCS Burden subscale loading invariant 2796.239 179 1 0.014 1 .91 ns <.0001 <.01

5. KDCS Social Interaction subscale loading invariant 2799.730 179 1 3.505 1 .062 ns .0004 .03

6. KDCS Cognitive subscale loading invariant 2796.928 179 1 0.703 1 .41 ns <.0001 .01

7. KDCS Effects subscale loading invariant 2798.687 179 1 2.462 1 .12 ns .0005 .03

8. KDCS Sleep subscale loading invariant 2811.091 179 1 14.866 1 .00012 .0036 .0003 .06

9. KDCS Social Support subscale loading invariant 2803.528 179 1 7.303 1 .0069 ns .0001 .04

10. Partially metric invariant model (factor loadings for KDCS Sleep & Social Support subscales noninvariant) 2810.567 189 1 14.342 11 .22 ns .0003 .06

11. Partially invariant model with 5 metric invariant KDCS subscale intercepts invariant 2894.471 194 10 83.904 5 .000001 .00005 .0019 .15

12. Partially invariant model with 8 metric invariant SF36 subscale intercepts invariant 2964.251 197 10 153.684 8 .000001 .00005 .0040 .21

13. Partially invariant model with intercept of KDCS Burden subscale invariant 2812.836 190 10 2.269 1 .14 ns .0003 .03

14. Partially invariant model with intercept of KDCS Social Interaction subscale invariant 2838.461 190 10 27.894 1 .000001 .00005 .0008 .09

15. Partially invariant model with intercept of KDCS Cognitive subscale invariant 2835.202 190 10 24.635 1 .000001 .00005 .0007 .08

16. Partially invariant model with intercept of KDCS Symptoms subscale invariant 2877.711 190 10 67.144 1 .000001 .00005 .0015 .14

17. Partially invariant model with intercept of KDCS Effects subscale invariant 2839.951 190 10 29.384 1 .000001 .00005 .0008 .09

18. Partially invariant model with intercept of SF-36 PF subscale invariant 2815.734 190 10 5.167 1 .024 ns .0004 .04

19. Partially invariant model with intercept of SF-36 RP subscale invariant 2846.345 190 10 35.778 1 .000001 .00005 .0001 .10

20. Partially invariant model with intercept of SF-36 BP subscale invariant 2819.639 190 10 9.072 1 .0026 ns .0004 .05

21. Partially invariant model with intercept of SF-36 GH subscale invariant 2810.568 190 10 0.001 1 .98 ns .0003 <.01

22. Partially invariant model with intercept of SF-36 MH subscale invariant 2837.769 190 10 27.202 1 .000001 .00005 .0008 .09

23. Partially invariant model with intercept of SF-36 RE subscale invariant 2900.352 190 10 89.785 1 .000001 .00005 .0018 .16

24. Partially invariant model with intercept of SF-36 SF subscale invariant 2831.587 190 10 21.020 1 .000005 .00016 .0007 .08

25. Partially invariant model with intercept of SF-36 VT subscale invariant 2810.914 190 10 0.347 1 .56 ns .0003 <.01

26. Partially metric invariant model with two-factor variances & covariance invariant 2816.786 192 10 6.219 3 .11 ns .0005 .04

27. Partially metric invariant model with factor variances-covariance & unique error variances for KDCS subscales invariant 2866.086 199 26 49.300 7 .000001 .00005 .0007 .12

28. Partially metric invariant model with factor variances-covariance & unique error variances for SF-36 subscales invariant 2840.570 200 26 23.784 8 .0025 ns <.0001 .09

29. Partially metric invariant model with factor variances-covariance & unique error variance for KDCS Burden subscale invariant 2827.202 193 26 10.416 1 .0013 .036 .0003 .07

30. Partially metric invariant model with factor variances-covariance & unique error variance for KDCS Social Interaction subscale invariant 2816.909 193 26 0.123 1 .73 ns .0006 .01

31. Partially metric invariant model with factor variances-covariance & unique error variance for KDCS Cognitive subscale invariant 2821.228 193 26 4.442 1 .036 ns .0001 .04

32. Partially metric invariant model with factor variances-covariance & unique error variance for KDCS Symptoms subscale invariant 2825.083 193 26 8.297 1 .004 ns .0001 .05

33. Partially metric invariant model with factor variances-covariance & unique error variance for KDCS Effects subscale invariant 2816.917 193 26 0.131 1 .72 ns .0006 .01

34. Partially metric invariant model with factor variances-covariance & unique error variance for KDCS Sleep subscale invariant 2838.330 193 26 21.544 1 .000004 .00013 <.0001 .08

35. Partially metric invariant model with factor variances-covariance & unique error variance for KDCS Social Support subscale invariant 2821.074 193 26 4.288 1 .039 ns .0009 .03

36. Partially metric invariant model with factor variances-covariance & unique error variance for SF-36 PF subscale invariant 2817.060 193 26 0.274 1 .61 ns .0006 .01

37. Partially metric invariant model with factor variances-covariance & unique error variance for SF-36 RP subscale invariant 2818.194 193 26 1.408 1 .24 ns .0004 .02

38. Partially metric invariant model with factor variances-covariance & unique error variance for SF-36 BP subscale invariant 2816.855 193 26 0.069 1 .80 ns .0007 <.01

39. Partially metric invariant model with factor variances-covariance & unique error variance for SF-36 GH subscale invariant 2819.464 193 26 2.678 1 .11 ns .0003 .03

40. Partially metric invariant model with factor variances-covariance & unique error variance for SF-36 MH subscale invariant 2817.791 193 26 1.005 1 .32 ns .0009 .02

41. Partially metric invariant model with factor variances-covariance & unique error variance for SF-36 RE subscale invariant 2821.873 193 26 5.087 1 .025 ns .0011 .09

42. Partially metric invariant model with factor variances-covariance & unique error variance for SF-36 SF subscale invariant 2821.253 193 26 4.467 .035 ns .0002 .04

43. Partially metric invariant model with factor variances-covariance & unique error variance for SF-36 VT subscale invariant 2826.729 193 26 9.943 .0017 .045 .0002 .05

Note: CFI = Comparative fit index. W2 = ratio of chi-square divided by N [68], which is analogous to R-squared (i.e., the proportion of explained variance) in multiple regression. Cohen [68] suggested that w2 ≤ 0.01 is small, w2 = 0.09 is medium, and w2 ≥ 0.25 is large.