In this paper we discuss the observational consequences of a dependence of meteor brightness on the entry angle z_R_, namely the decrease of the observed rates of meteors and the reduction of the population indices r for radiant zenith distances larger than zero. These effects, if not properly accounted for, lead to erroneous values of meteor shower activity. The first one is corrected by applying a factor cos^-γ^ z_R_ to the observed rates, while the second one requires a correction of the form r(0deg)=r(z_R_)^b(z_R_)/b(0deg)^. We give general expressions for γ and b(z_R_)/b(0deg) in terms of the coefficients relating the absolute magnitude of the meteor to its physical parameters. High quality photographic data, as well as naked-eye observations, are analyzed. It turns out that γ=1+1.53logr(0deg) describes the photographic behavior quite well. For visual observations, however, an exponent γ=1 should be employed in most cases. We explain this difference on the basis of a larger path of the meteoroids at increasing entry angles, which increases the probability of perception of meteors for visual observers. As regards b(z_R_)/b(0deg), a dependence b(z_R_)/b(0deg)=1.04-0.04cos z_R_ is suitable for both photographic and naked-eye records. These values are compared to predictions resulting from a detailed treatment of the single body theory. It turns out that the single body approach gives a much stronger dependence of the meteor brightness on the entry angle than that obtained from observations. Models including quasicontinuous fragmentation are therefore necessary.