|
|
| Message/Author |
|
|
|
|
I've been running SEM models that have strange model fit indices.
CFI=.701
TLI=.688
RMSEA=.042
Does anyone know what it means when the RMSEA fit is acceptable, but the CFI and TLI values are poor? |
|
|
|
|
| I don't think there is any general guidance on this. One possible explanation is that the variables have low correlations so that the baseline model of CFI (uncorrelated variables) doesn't fit too poorly and therefore the H0 model doesn't fit that much better. Low RMSEA and low chi-square also goes with low correlations. |
|
|
| Back to top |
|
|
