Page 92 - Ana Kozina and Nora Wiium, eds. ▪︎ Positive Youth Development in Contexts. Ljubljana: Educational Research Institute, 2021. Digital Library, Dissertationes (Scientific Monographs), 42.
P. 92
positive youth development in contexts
Measurement invariance
Table 2 shows fit indexes of MGCFA of the DA. For gender, the configural
invariance model indicated an adequate fit, meaning that similar patterns
of observed and latent constructs across gender emerged. In the metric
invariance model, in which factor loadings on latent variables were con-
strained to be equal across gender, the fit indices showed construct and
metric equivalence across countries. In the scalar invariance model, in
which variables were fixed to have the same intercept across gender, an ad-
equate or good model fit was not achieved. Following the modification in-
dices (means of items 33 and 31 were allowed to vary across gender), partial
scalar invariance was attained. As for school level, the configural invari-
ance model, metric invariance model and scalar invariance model indicat-
ed an adequate fit, meaning scalar invariance for school level of the DA was
achieved.
Table 2. Measurement invariance models and goodness-of-fit indexes of multigroup analyses
of developmental assets across gender and school types
Model Model fit indices
Gender χ2 (df) RMSEA 90% CI RMSEA CFI
Configural invariance
Metric invariance 6333.25 (2840) .035 .034–.036 .915
Scalar invariance 7074.65 (3272) .034 .033–.035 .908
Partial scalar 7518.88 (3326) .036 .035–.037 .898
invariance 7424.97 (3324) .900
.035 .034–.036
School type 6349.89 (2840) .917
Configural invariance 7077.40 (3272) .035 .035–.036 .910
Metric invariance 7289.87 (3326) .034 .033–.035 .906
Scalar invariance .035 .034–.036
Notes. χ2 = Chi-square; df = degrees of freedom; CFI – Comparative Fit Index; RMSEA = root mean
square error of approximation; CI = confidence interval
In Table 3, fit indexes of MGCFA of the 5Cs are presented. For gender,
the configural invariance and metric invariance model indicated an ade-
quate fit. In the scalar invariance model, in which variables were fixed to
have the same intercept across gender, an adequate or good model fit was
not achieved. Following the modification indices (means of item 33 were al-
lowed to vary across gender and items 22 and 23 were allowed to correlate
since it was justified by the content), partial scalar invariance was attained.
92
Measurement invariance
Table 2 shows fit indexes of MGCFA of the DA. For gender, the configural
invariance model indicated an adequate fit, meaning that similar patterns
of observed and latent constructs across gender emerged. In the metric
invariance model, in which factor loadings on latent variables were con-
strained to be equal across gender, the fit indices showed construct and
metric equivalence across countries. In the scalar invariance model, in
which variables were fixed to have the same intercept across gender, an ad-
equate or good model fit was not achieved. Following the modification in-
dices (means of items 33 and 31 were allowed to vary across gender), partial
scalar invariance was attained. As for school level, the configural invari-
ance model, metric invariance model and scalar invariance model indicat-
ed an adequate fit, meaning scalar invariance for school level of the DA was
achieved.
Table 2. Measurement invariance models and goodness-of-fit indexes of multigroup analyses
of developmental assets across gender and school types
Model Model fit indices
Gender χ2 (df) RMSEA 90% CI RMSEA CFI
Configural invariance
Metric invariance 6333.25 (2840) .035 .034–.036 .915
Scalar invariance 7074.65 (3272) .034 .033–.035 .908
Partial scalar 7518.88 (3326) .036 .035–.037 .898
invariance 7424.97 (3324) .900
.035 .034–.036
School type 6349.89 (2840) .917
Configural invariance 7077.40 (3272) .035 .035–.036 .910
Metric invariance 7289.87 (3326) .034 .033–.035 .906
Scalar invariance .035 .034–.036
Notes. χ2 = Chi-square; df = degrees of freedom; CFI – Comparative Fit Index; RMSEA = root mean
square error of approximation; CI = confidence interval
In Table 3, fit indexes of MGCFA of the 5Cs are presented. For gender,
the configural invariance and metric invariance model indicated an ade-
quate fit. In the scalar invariance model, in which variables were fixed to
have the same intercept across gender, an adequate or good model fit was
not achieved. Following the modification indices (means of item 33 were al-
lowed to vary across gender and items 22 and 23 were allowed to correlate
since it was justified by the content), partial scalar invariance was attained.
92
