The spaced antenna and interferometry techniques are applied to data collected on two MF radars (at Christchurch, New Zealand, and at Scott Base, in the Antarctic) and one VHF radar (the MU radar at Shigaraki, Japan). Velocity results obtained by the two techniques are compared as functions of characteristics such as the temporal scale and anisotropy of the scatter. In general, there was good agreement between the two techniques; however, when the temporal scale is small (indicating turbulent scatter) there was less agreement. The interferometry technique did not appear to be influenced greatly by anisotropy in the scatter. The interferometry technique was found to be affected by averaging of the Doppler spectra and by the use of an extra receiver. It is suggested that less averaging of the spectra results in increased random noise components in the spectra, which in the phase spectra are interpreted as scattering from a greater range of zenith angles. It was found that the use of three receivers rather than four also lead to scattering that appeared to originate from a greater range of zenith angles. These effects lead to lower interferometric velocity estimates that are more biased towards full correlation analysis (FCA) true velocities.