Abstract
We discuss various aspects of self-energy corrections to the fermion propagator at a finite temperature. Several calculational methods not relying on the renormalization of the mass or of the wave function are discussed. General expressions are given. It is shown that, when calculating a physical process, the thermal mass enters essentially the phase-space factor and not the dynamical part of the process, at least in some limiting cases. Comparison with the renormalization approach is discussed and it appears that there are no ambiguities in the calculation of the self-energy corrections at finite temperature.