It is shown that in the measurement of the aspect sensitivity of radio aurora the errors arising in the determination of the aspect of the angles are as important as the errors arising in the determination of the backscatter cross section. The former errors have not been adequately assessed in most of the published measurements.