A new method of evaluating relative (arm with respect to the interarm disk) star formation rates and relative star formation efficiencies, together with spiral arm-amplitudes, as a function of the galactocentric radius, using broad-band photometry is derived. The classical method for obtaining star formation rates from H_alpha photometry is discussed, and a new method is derived for diagnosing the possible presence of biased star formation due to different initial mass functions in the arms and in the interarm disk. As an example, these methods are applied to the spiral galaxies NGC 4321 and NGC 4254, obtaining their arm amplitudes, relative arm/interarm star formation efficiencies, and relative arm/interarm initial mass functions for each arm, as a function of the galactocentric radius. Both objects present evidence of massive star formation triggering in the spiral arms consistent with the spiral density-wave theory, and a different initial mass function in the arms from that in the rest of the disk, in the sense of favoring a larger fraction of massive stars in the arms. This biased star formation is present in the zones of the arms where there is triggering of massive star formation, and is then related to, and probably caused, by the density-wave system. However, due to this biased star formation, the spiral arms of the galaxies studied do not trigger the formation of a larger total mass of stars (of all spectral types) with respect to the interarm disk.