Why does the GFI have a negative value?

My SEM model has 3 observed variables.

Two are independent variables and one is the dependent variable.

Can someone please tell me, it is urgent.

Hi ​,

Without seeing your model I can only guess at the cause or encourage you to explore Google for the theory underlying Goodness of Fit indices. Ultimately, given you are getting results and no errors or warnings, this will come down you diagnosing your model. Something about the model you specified (the implied covariance matrix) doesn't match well with the underlying observed covariance matrix.

My guesses:

Is your model identified?

Do the independent variables covary (double-headed curved arrow).

Do the arrow heads from the Independent variables point at the Dependent variable?

Does the Dependent exogenous variable have an exogenous unobserved error variable? Is the variance on that arrow fixed at 1?

Are there other errors or warnings the Amos user interface is giving you?

I hope this helps,

​