A "gradient correction factor" and an "electron fluence correction factor" are used in the AAPM Task Group 21 dosimetry protocol, the product of the two being the "replacement correction factor." The separation of two physical processes, implied by the concepts, is theoretically unfounded. The proof of this statement rests on Fano's theorem, the conditions for which constitute a case without variations in fluence or its directional and energy distributions. Deviations from these conditions represent gradients in the radiation field, affect the response of a measurement cavity, and do so by perturbing the fluence of electrons that deposit energy in the cavity. Thus, in beams of both electrons and photons, corrections are required when fluence variations exist in the vicinity of the cavity. The rationale for an electron fluence correction based on "in-scattering" and "obliquity" is weak, since the effects only occur when gradients are present. The choice of values for the two correction factors at the depth of maximum dose are particularly inconsistent with the actual characteristics of the radiation field.