Abstract: From the RHIC/PHENIX experiment data, it is found that the statistical error of spin transfer is few times larger than the statistical error of the single spin asymmetry. In order to verify the difference between σD_LLand σA_L, the linear least-squares method was used to check it first, and then a simple Monte-Carlo simulation to test this factor again. The simulation is consistent with the calculation result which indicates that the few times difference is reasonable.