A note on local identifiability and degrees of freedom in the asymptotic likelihood ratio test

It is pointed out that both the degrees of freedom in the chi-square asymptotic likelihood ratio test and the local identifiability of the parameters in the model can be determined unambiguously from the ranks of certain matrices.