For the determination of the experimental value of the anisotropy field we use the value of the saturation field deduced from the hysteresis loop...what is the reason ?
There is a very nice and thorough explanation of this measurement given in section 7.6 of Introduction to Magnetic Materials, 2nd edition by Cullity and Graham. ISBN 978-0471-447741-9. I think this classic text may address your question.
I agree that the most clear explanation on the different methods is on the reference given above by S Trout (Cullity). HOwever nowadays the syngle-crystal required for these measurements are difficult to obtain, so alternative techniques are used to find K1 and K2 and from there go to the anisotropy field.
Really, this is not always so. And even in crystals with uniaxial magnetic anisotropy, it must also consider the shape of the sample (there is the so-called shape anisotropy associated with the demagnetizing fields). The concept of effective anisotropy field is introduced as the magnetic field, which near easy magnetization axis (EMA) acts on the magnetic moment of the material similar to that of anisotropy (the energy of interaction with such field is equal to the energy of anisotropy at small deviations from EMA). In the case of pure uniaxial anisotropy Hk = 2Ku/Ms = Hs. In thin films with perpendicular anisotropy we have: 2Ku/Ms – 4πMs =< Hs < 2Ku/Ms, where Hs is the saturation field in the plane of the film, 4πMs is demagnetizing field in Gaussian system of units. In magnetically multiaxial crystals, depending on the sign of the anisotropy constant K1, the anisotropy field, for example, can be equal to 2K1/Ms or 4K1/3Ms and depends on the direction of deviation of the magnetization vector from EMA.