The present work describes a study about the influence of anisotropy in a RANS based prediction scheme for noise prediction of subsonic single jet flows. This task was motivated by an ongoing research to improve the spectral predictions obtained by using Lighthill Acoustic Analogy. The numerical approach employed is originally based on the MGBK technique. Nevertheless, an improved energy-transfer time scale (TET) has been utilized in order to enhance the spectrum shape prediction. In the present model, the anisotropy effect on the noise prediction is indirectly coupled with the aerodynamic flow calculation. However, in order to unveil the actual contribution of the anisotropy, the coupling with the aerodynamic calculation has been performed through the computation of turbulence quantities as the stress tensor components by using a Reynolds Stress Turbulence Model (RSTM). The numerical results permit to identify the effect of the anisotropy parameters and its effect in the final spectrum shape prediction. This approach has indicated some limitations and some potential ways to improve the current methodology towards jet noise prediction of subsonic single flows.