Skip to main content

Table 2 Confirmatory factor analysis: Model comparison and tests of measurement invariance for the nine-factor model (Robust Maximum Likelihood estimation - MLM)

From: Factor structure and psychometric properties of the trier inventory for chronic stress (TICS) in a representative german sample

Model tested Absolute fit Comparison to less-constrained (or baseline) model Comparative fit Parsimony fit
  N c1 SB χ2 df SB χ2/df Δdf T S (ΔSB χ2) AIC TLI CFI SRMR RMSEA a
Factorial Invariance             
Two-factor model 1,149 1.259 11,847.949*** 1,538 7.7    157,955 .747 .756 .078 .067
Nine-factor model 1,149 1.250 7,383.383*** 1,500 4.9 38 2,765.7*** 153,564 .855 .863 .071 .051
Measurement Invariance (Gender)             
Female 622 1.209 4,837.946*** 1,500 3.2    82,414 .858 .867 .071 .052
Male 527 1.220 4,720.306*** 1,500 3.1    71,065 .830 .840 .077 .055
Baseline Model 1,149 1.211 9,665.630*** 3,048 3.2    153,488 .846 .853 .075 .053
Partial measurement invariance b 1,149 1.209 9,761.882*** 3,099 3.2 51 88.4** 153,482 .847 .852 .077 .053
Full measurement invariance c 1,149 1.210 9,874.099*** 3,157 3.1 58
(109)
88.8***
(176.4***)
153,482 .849 .851 .078 .053
Measurement Invariance (Age)             
Young (14-47) 540 1.193 4,562.580*** 1,500 3.0    73,432 .834 .844 .075 .054
Old (48-99) 609 1.251 4,743.958*** 1,500 3.2    79,730 .863 .871 .068 .050
Baseline Model 1,149 1.219 9,481.756*** 3,048 3.1    153,245 .849 .856 .073 .052
Partial measurement invariance b 1,149 1.215 9,571.958*** 3,099 3.1 51 92.4*** 153,236 .850 .854 .075 .052
Full measurement invariance c 1,149 1.217 9,812.218*** 3,157 3.1 58
(109)
181.5***
(248.6***)
153,358 .849 .850 .076 .052
  1. Measurement invariance tested for the nine-factor structure. c1 = χ2 scale correction factor for Robust Maximum Likelihood estimation (MLM). SB χ2 = Satorra-Bentler scaled χ2 value. T S (ΔSB χ2) = test statistic for the scaled difference in SB χ2s (SDCS test) with df as stated above comparing a model with the model above. ** p < .01, *** p < .001. AIC = Akaike information criterion. a RMSEA 90% confidence interval is not available for MLM estimations in MPlus 5.1. b Factor loadings and indicator intercepts constrained as equal. C Factor loadings, indicator intercepts, and indicator errors hold equal