What we know from looking at the experimental spectra is the fact that the anisotropy measure becomes smaller for larger external magnetic fields.
The splitting between the energy levels depends on the size of the external magnetic field. From the spectra, we can see that when this external magnetic field is larger, the frequency becomes smaller. Yet, from \Delta E ~ B_0 you would expect the energy splittings to become larger, I think? So, from this reversed result I would think the magnetic hyperfine field to be in the inverse direction than that of the applied magnetic field?