Skip to main content
. Author manuscript; available in PMC: 2020 Jul 7.
Published in final edited form as: Child Psychiatry Hum Dev. 2019 Jun;50(3):473–482. doi: 10.1007/s10578-018-0854-0

Table 6:

Invariance Tests by Age Group

Child Report
Invariance Level χ2 df Δ χ2 CFI RMSEA Lower CI Upper CI Change in CFI Change in RMSEA
Configural 3286.088 1538 -- 0.967 0.046 0.044 0.048 -- --
Metric 2962.415 1579 54.17 0.974 0.040 0.038 0.042 0.007 −0.006
Threshold 3073.168 1661 159.77*** 0.974 0.040 0.037 0.042 0 0
Residual 3272.629 1610 97.50*** 0.969 0.044 0.041 0.046 −0.005 0.004
Parent Report
Invariance Level χ2 df Δ χ2 CFI RMSEA Lower CI Upper CI Change in CFI Change in RMSEA
Configural 2780.89 1538 -- 0.954 0.039 0.036 0.041 -- --
Metric 2728.21 1579 103.82*** 0.958 0.037 0.034 0.039 0.004 −0.002
Threshold 3159.85 1661 168.31*** 0.945 0.041 0.039 0.043 −0.013 0.004
Residual 3159.85 1660 91.42*** 0.946 0.041 0.039 0.043 .001 0

Invariance tests using a mean- and variance-adjusted weight least squares (WLSMV) estimator comparing older and younger children on child-report (top panel) and parent-report (bottom panel) are presented here. Good model fit (noted in bold) was established by a comparative fit index (CFI) > .95 and root mean square error of approximation (RMSEA) < .06. Measurement invariance at each level (noted in bold) was established by a decrement in CFI < .01 and an increase in RMSEA < .015.

***

p < .001.