Skip to main content
. 2019 Nov 18;14(11):e0222277. doi: 10.1371/journal.pone.0222277

Table 4. Analysis of factorial invariance for age and gender using multigroup confirmatory factor analyses.

Χ2scaled df CFI ΔCFI RMSEA ΔRMSEA Measurement Invariance Testa
Gender
(group 1 = men; group 2 = women)
Model 0 configural invariance 190.6 54 .976 .057
Model 1 weak invariance 203.8 62 .976 +.000 .053 +.004
Model 2 strong invariance 249.0 70 .971 +.005 .055 -.002
Model 3 strict invariance 252.3 79 .971 +.000 .051 +.004
Age
(group 1: < 25 y; group 2: 25 to 34 y; group 3: 35 to 44 y; group 4: 45 to 54 y; group 5: 55 to 64 y; group 6: 65 to 74 y; group 7: ≥ 75 y)
Model 0 configural invariance 347.7 189 .970 .063
Model 1 weak invariance 412.6 237 .969 +.001 .057 -.006
Model 2 strong invariance 588.8 285 .951 +.018 .066 -.009
Model 3 strict invariance 741.2 339 .935 +.016 .069 -.004

Notes: df = degrees of freedom; CFI = robust version of the Comparative Fit Index; ΔCFI = differences between models (0 and 1, 1 and 2; 2 and 3) regarding robust CFI; RMSEA = robust version of the root mean square of approximation; ΔRMSEA = differences between models (0 and 1, 1 and 2; 2 and 3) regarding robust RMSEA

a = ΔCFI ≤ .010 supplemented by ΔRMSEA ≥ -.015 indicates non-invariance. √ marks invariance