A statistical analysis of the elastic unpolarized electron proton scattering data shows that, at large momentum transfer, the size and the $\\epsilon$ dependence of the radiative corrections, as traditionally calculated and applied, may induce large correlations of the parameters of the Rosenbluth fit, which prevent a correct extraction of the electric proton form factor. Using the electron QED structure (radiation) function approach the cross section of elastic electron-proton scattering in leading and next-to leading approximations is calculated and expressed as a correction to the Born cross section, which is different for the electric and the magnetic contribution. When properly applied to the data, it may give the solution to the problem of the discrepancy of the polarized and unpolarized results on electron proton scattering. |
