Table 4. Summary table measurement invariance analysis (n = 222 532 in 44 countries).
Model fit | Change in model fit | ||||||
---|---|---|---|---|---|---|---|
Par. | CFI | TLI | RMSEA | SRMR | ΔCFI | ΔRMSEA | |
Country invariance | |||||||
Configural | 792 | 0.979 | 0.972 | 0.037 | 0.040 | ||
Scalar | 491 | 0.969 | 0.967 | 0.040 | 0.045 | −0.010 | 0.003 |
Gender invariance | |||||||
Configural | 36 | 0.979 | 0.972 | 0.035 | 0.034 | ||
Scalar | 29 | 0.978 | 0.974 | 0.034 | 0.034 | −0.001 | −0.001 |
By country, minimum | −0.006 | ||||||
By country, maximum | 0.003 | ||||||
Age invariancea | |||||||
Configural | 54 | 0.975 | 0.967 | 0.035 | 0.034 | ||
Scalar | 40 | 0.974 | 0.970 | 0.033 | 0.034 | −0.001 | −0.002 |
By country, minimum | −0.013 | ||||||
By country, maximum | 0.008 | ||||||
Socio-economic invarianceb | |||||||
Configural | 54 | 0.981 | 0.975 | 0.035 | 0.033 | ||
Scalar | 40 | 0.982 | 0.979 | 0.032 | 0.033 | 0.001 | −0.003 |
By country, minimum | −0.002 | ||||||
By country, maximum | −0.001 |
Par. = number of free parameters; CFI = comparative fit index; TLI = Tucker Lewis index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual.
n = 221 093 due to missing values of age;
n = 212 353 due to missing values of socio-economic status.