The theoretical models for amplitude-dependent damping are worked out for hypothetical samples under homogeneous strains. This is called the intrinsic damping. Experimentally, however, a strain distribution is imposed onto the sample and the measured data must be corrected, for a comparison with the theoretical models. This problem has been treated in the internal friction literature, for the case where the spatial strain distribution in the specimens remains constant during each experiment, i.e., when the internal friction is measured at constant maximum strain amplitude. The situation is more complicated when the internal friction is amplitudedependent and the data are obtained in free decay, so that, in addition to a spatial strain distribution, the maximum strain amplitude is changing during each vibration period. This problem has not been considered in the literature or the correction due to the strain distribution has been wrongly applied to internal friction data, obtained during free decay, by taking into account only the distribution in space and ignoring the fact that the strain amplitude is also time dependent. Since several amplitude-dependent internal friction data, taken during free decay, are reported in the literature, and several dislocation parameters are obtained from these results, it is important to evaluatemore » the influence of variable strains (in space and time) on the shape of the intrinsic damping versus strain curves.« less