Skip to main content
. 2016 Sep 22;11(9):e0162510. doi: 10.1371/journal.pone.0162510

Table 3. Goodness-of-fit for the DEBQ three-factor model across different groups and measurement invariance across gender, age and BMI.

CFI TLI RMSEA (90%CI) SRMR Δ CFI
Split half sample (N2 = 1205) Three-factor model .903 .895 .068 (.066-.071) .054
Three-factor model (a) .937 .931 .055 (.053-.058) .054
Gendera
Women (N = 1052) .930 .924 .059 (.056-.061) .057
Men (N = 1305) .920 .913 .061 (.058-.064) .057
Multi-group analysisa
Model 0 (configural invariance) .926 .919 .042 (.041-.044) .057
Model 1 (metric invariance) .925 .921 .042 (.040-.043) .059 .001
Model 2 (scalar invariance) .917 .916 .043 (.042-.044) .060 .009
Agea
14–24 years (N = 258) .893 .884 .074 (.068-.080) .068
25–34 years (N = 360) .914 .907 .065 (.060-.070) .063
35–44 years (N = 348) .928 .922 .062 (.057-.067) .061
45–54 years (N = 438) .903 .895 .070 (.066-.074) .065
55–64 years (N = 435) .924 .918 .060 (.056-.065) .056
65–74 years (N = 328) .913 .905 .066 (.061-.072) .060
≥ 75 years (N = 190) .890 .881 .069 (.062-.077) .074
Multi-group analysisa
Model 0 (configural invariance) .904 .906 .025 (.024-.025) .096
Model 1 (metric invariance) .903 .906 .025 (.024-.025) .095 .001
Model 2 (scalar invariance) .898 .907 .025 (.024-.025) .095 .006
BMIa
Underweight/normal weight (N = 1038) .929 .923 .057 (.055-.060) .057
Overweight (N = 912) .919 .912 .063 (.060-.066) .055
Obesity (N = 369) .912 .905 .066 (.062-.071) .071
Multi-group analysisa
Model 0 (configural invariance) .922 .916 .035 (.034-.036) .057
Model 1 (metric invariance) .921 .918 .035 (.034-.036) .058 .001
Model 2 (scalar invariance) .915 .916 .035 (.034-.036) .058 .007

Notes. CFA = confirmatory factor analysis; CFI = comparative fit index; CI, confidence interval, RMSEA = Root Mean Square Error of Approximation; SRMR = standardized root mean square residual; three-factor model = correlated factors with cross-loadings fixed to 0; three-factor model (a) = correlated factors with cross-loadings fixed to 0 and correlation of unique variances between items 8 and 9 (diffuse emotions); Δ CFI, difference values between model 1/2 and model 0

a, three-factor model with correlation of unique variances between items 8 and 9 (diffuse emotions).