Identification strength - where did the yellow bar go?

Hi there,
I have sth that I do not understand concerning my identification check.
What does it mean when a parameter seems well identified relative to param value but unidentified relative to the prior standard deviation?
See attached the standard yellow/blue bar plot. I’ve read the Pfeifer manual but that case of no yellow bar was not covered :frowning:
Any help is greatly appreciated!

identi.pdf (5.8 KB)

In newer Matlab versions the coloring has changed, but the legend should still the same. The difference between the two curves refers to the normalization of the curvature of the likelihood function. One cannot divide by the parameter value if that value is (close to) 0. But looking at your graph, there seem to be no serious problems.