This is because Raman scattering is an indirect probe of stress. The stress produces changes in the interatomic interactions that lead to the frequency shifts. But in order to determine the stress one must have complete information about the dynamics and Raman scattering gives only the zero wavevector frequencies. Even with the complete information about the dynamics (from inelastic neutron or X-ray scattering) it is not always easy to get the interatomic fores or potentials because one must have an appropriate model as well and/or ab intitio calculations. X-ray diffraction however gives direct information about the shifted atom positions in the thin films compared to the bulk from the information of the peak positions and intensities.
Although Raman scattering can give indirect information about the stress, X-ray diffraction experiments can give direct and hence more reliable measurements of the stress in thin films.
This is because Raman scattering is an indirect probe of stress. The stress produces changes in the interatomic interactions that lead to the frequency shifts. But in order to determine the stress one must have complete information about the dynamics and Raman scattering gives only the zero wavevector frequencies. Even with the complete information about the dynamics (from inelastic neutron or X-ray scattering) it is not always easy to get the interatomic fores or potentials because one must have an appropriate model as well and/or ab intitio calculations. X-ray diffraction however gives direct information about the shifted atom positions in the thin films compared to the bulk from the information of the peak positions and intensities.
I agree with the previous post (Chatterji), but with the caveat that you need to consider the volume being averaged with each technique. For large changes in stress over short length scales, the volume averaging issue can be very significant, and may be part of the reason for a discrepancy between Raman and XRD, for example.
Well one can focus at least synchrotron x-rays to a few hundredths of a nanometer or even better. I guess with laser light one can do so easily on the thin film. So you can look at the film with different averaging area and volume. If you can keep the volume in both techniques same, then you may get consistent results. I guess the problem lies in the differences in two techniques. I shall consider X-ray results to be more reliable because it is direct.