Skip to main content
. 2020 Mar 2;20:51. doi: 10.1186/s12874-020-00927-x

Table 3.

Analysis of measurement invariance

Model χ2(df) Δχ2 p CFI ΔCFI GH ΔGH
Gender multi-group analysis (female / male)
 Configural invariance 248.92 (26) .967 .951
 Metric invariance 265.85 (31) 16.93 .342 .965 .002 .949 .002
 Scalar invariance 278.93 (36) 13.08 < .001 .964 .001 .947 .002
 Strict invariance 269.00 (43) 9.93 .248 .967 .003 .951 .004
Age multi-group analysis (≤39, 40–59, ≥60)
 Configural invariance 250.44 (39) .970 .932
 Metric invariance 274.31 (49) 23.87 .008 .968 .002 .927 .005
 Scalar invariance 409.34 (59) 135.03 < .001 .950 .018 .891 .036
 Partial scalar invariance a 298.23 (53) 23.92 < .001 .965 .003 .922 .005
 Partial strict invariance a 294.17 (67) 4.06 .852 .967 .002 .927 .005
Education multi-group analysis (≥ 9 years, 10 years, ≤ 11 years)
 Configural invariance 249.921 (39) .973 .976
 Metric invariance 272.582 (49) 22.661 .012 .973 .000 .974 .002
 Scalar invariance 335.692 (59) 63.110 < .001 .967 .006 .968 .006
 Strict invariance 332.419 (73) 3.273 .998 .967 .000 .970 .002
Family status multi-group analysis (married, single, divorced, widowed)
 Configural invariance 241.489 (52) .973 .976
 Metric invariance 288.811 (67) 47.322 < .001 .972 .001 .973 .003
 Scalar invariance 378.669 (82) 89.858 < .001 .965 .007 .964 .009
 Strict invariance 415.905 (103) 37.236 .016 .960 .005 .962 .002
Employment status multi-group analysis (working, retired, school/apprenticeship, not working)
 Configural invariance 273.037 (52) .974 .975
 Metric invariance 306.536 (67) 33.499 .004 .974 .000 .973 .002
 Scalar invariance 457.330 (82) 150.794 < .001 .960 .014 .959 .014
 Partial scalar invarianceb 402.601 (79) 96.065 < .001 .965 .009 .964 .009
 Partial strict invarianceb 422.636 (100) 20.035 .331 .963 .002 .964 .000

aIntercepts of Items 1, 2, and 4 were freed to vary between groups in these models

bThe intercept of Item 2 was freed to vary between groups in these models