Skip to main content
. 2022 Dec 22:1–23. Online ahead of print. doi: 10.1007/s10869-022-09869-1

Table 2.

Multi-level confirmatory factor analyses (MCFAs) for measurement models

Time 1 Time 2 Time 3 Time 4 Time 5 Time 6
A. Models based on five constructs: safety climate, emotional exhaustion, safety compliance, morale, and depression
χ2 1024.10 744.80 868.44 889.31 936.85 728.84
df 381 381 381 381 381 381
CFI .874 .911 .894 .905 .885 .902
RMSEA .051 .037 .043 .043 .045 .036
SRMR within .056 .059 .063 .057 .061 .064
SRMR between .085 .101 .062 .038 .048 .058
r (emotional exhaustion and depression) .50 .51 .47 .46 .44 .41
B. Models in which emotional exhaustion and depression were merged
χ2 1208.38 992.17 1008.68 1105.70 1148.98 957.13
df 385 385 385 385 385 385
CFI .840 .851 .864 .865 .843 .839
RMSEA .057 .047 .048 .051 .053 .046
SRMR within .065 .068 .075 .068 .083 .085
SRMR between .085 .101 .062 .038 .048 .058
A–B Model comparison
Δχ2 184.28 247.37 140.24 216.39 212.13 228.29
Δ CFI .03 .06 .03 .04 .04 .06

All Δχ2 were statistically significant (p < .01) and all Δ CFI values were greater than .02 (Cheung & Rensvold, 2002), suggesting significant model fit deterioration after the merge of the emotional exhaustion and depression constructs