[en] Inelastic M3 form factors for the excitation of the 5.236 MeV 3⁺ state in ²⁴Mg and for the excitation of the 10.376 MeV 3⁺, T = 1 state in ²⁸Si have been measured. These form factors are compared with full 2s1d shell-model calculations in order to investigate the experimental quenching of M3 transition strength which is expected in the light of previous work. The shell-model predictions do not agree with the experimental data. In the case of the ²⁴Mg 5.236 MeV state the data are enhanced by a factor of approx.= 2 with respect to the shell-model prediction, whereas in the case of the ²⁸Si 10.376 MeV state the data are suppressed by a factor of approx.= 5. In the former case it is found that a small admixture of the 3₂⁺ state shell-model wave function into the 3₁⁺ state wave function gives good agreement with the data. The effects of this admixture on other observables are investigated. No similar mechanism is found to explain the ²⁸Si M3 transition. In view of the basic disagreement between the shell-model predictions and the experimental data, no conclusions about M3 quenching can be drawn. The validity of previous analyses using this technique is seriously questioned. (orig.)