Normally, test reliability in CTT is about alpha coefficient. But, in IRT it is about 'test information function (TIF)'.

Let say by applying CTT in a test analysis, we have an alpha coefficient reliability result of 0.75. Then, we can easily refer this result to a kind of 'interpretation ranges' to make a judgement for the test reliability; e.g. poor, acceptable, high etc.

But, for TIF in IRT application, do we have any 'interpretation ranges' like alpha coefficient reliability? Let say I have a test with 40 item and TIF = 499.92, how am I going to say about the test reliability?

Perhaps somebody that familiar with IRT can help me here.