Numerical simulations are an important tool in furthering our understanding of turbulent dynamo action, a process that occurs in a vast range of astrophysical bodies. It is important in all computational work that comparisons are made between different codes and, if non-trivial differences arise, that these are explained. In a recent paper, Käpylä, Korpi & Brandenburg describe an attempt to reproduce some of our results and, by employing a different methodology, they arrive at very different conclusions concerning the mean electromotive force and the generation of large-scale fields. Here we describe why the simulations of Käpylä et al. are simply not suitable for a meaningful comparison, since they solve different equations, at different parameter values and with different boundary conditions. Furthermore, we describe why the ‘resetting’ method adopted by Käpylä et al. to calculate the α-effect is inappropriate, since the resulting value of α cannot be related to the evolution of any large-scale magnetic field.