Skip to main content
. Author manuscript; available in PMC: 2017 Apr 1.
Published in final edited form as: Psychol Assess. 2015 Aug 24;28(4):429–442. doi: 10.1037/pas0000186

Table 5. Results of Invariance Tests for the Revised Model.

Age Invariance χ2 df CFI RMSEA lower CI upper CI RMSEA p Change in CFI Change in RMSEA
Configural 635.931 255 0.978 0.058 0.053 0.064 0.009 - -
Weak 733.768 291 0.974 0.059 0.053 0.064 0.003 -0.004 0.001
Strong 977.839 327 0.962 0.067 0.062 0.072 0.000 -0.012 0.008
Strict 977.839 375 0.965 0.060 0.056 0.065 0.000 0.003 -0.007

Sex Invariance χ2 df CFI RMSEA lower CI upper CI RMSEA p Change in CFI Change in RMSEA

Configural 338.317 102 0.986 0.046 0.040 0.051 0.894 - -
Weak 351.180 111 0.986 0.044 0.039 0.050 0.963 -0.002 -0.001
Strong 398.220 120 0.984 0.046 0.041 0.051 0.912 0.002 0.002
Strict 398.220 132 0.984 0.043 0.038 0.048 0.993 -0.003 -0.003

Fit statistics for models testing age invariance are on the top half of the table; models testing sex invariance are on the bottom half of the table. The configural invariance models included all age groups or both the male and female groups in one model. The weak invariance model fixed factor loadings to be equal across all groups. The strong invariance model fixed factor loadings and item thresholds to be equal across groups. The strict invariance model fixed factor loadings, item thresholds, and unique/residual item variances to be equal across groups. Change in CFI and RMSEA were calculated for each model minus the previous model; negative change in CFI and positive change in RMSEA indicate a decrement in fit. Fit indices and change in fit indices passing criteria for good fit are in bold (CFI ≥ 0.95; RMSEA ≤ 0.06, RMSEA upper CI ≤ 0.08; RMSEA p>0.05; change in CFI ≥ -0.01; change in RMSEA ≤ 0.015).