I believe that you are referring to the FRF estimation method, which is not technically part of the modal parameter identification - but of course can have an influence on those results. Per the Help documentation for Spectral Acquisition, there are three choices available:
You can select the method from the dropdown list for estimating the frequency response function.
Most commonly-used method. It assumes no signal noise on the input signals and so no input errors. This provides a better estimate of the anti-resonances than of the resonances.
This is the best method overall but takes the longest processing time. The signal noise can be divided between input and output signals. It provides a good estimate of both the anti-resonances and the resonances.
This estimation method assumes no noise on the output signals so no output errors. It provides a better estimate of the resonances than of the anti-resonances.
The selection only becomes sensitive if the function in the upper-left is set to FRF. Then you will also have to pick your reference channel, etc. In your screen shot the function is set to a time block, and it looks like you are using Signature Testing as well. Do you have access to Spectral Testing, Impact Testing, MIMO FRF, etc.? You can measure FRF's in Signature, but it takes a bit more effort from the user than in the other workbooks designed for structural vibration data acquisition.