The electrical response of dielectric materials is often graphically represented by means of a plot in the complex impedance (or admittance) plane. In many cases these impedance plots are semicircles. An equivalent circuit consisting of a capacitor (or a constant phase element) and two resistors is often used to describe such a semicircle. It is shown here that even if the experimental data are lying on such a semicircle they do not necessarily correspond to that equivalent circuit because of a different frequency dependency. In addition it is shown that the Kramers-Kronig transforms can be very helpful but are not decisive for the determination of whether experimental errors are made.