The β-spectrum of RaE is reinvestigated taking into account the finite de Broglie wave length effect pointed out by Rose et al. and the correction to the nuclear matrix elements pointed out by the author, and it is shown that the conclusion of Petschek and Marshak that only the assumption of tensor+pseudoscalar, spin change 0-0, parity change yes is able to explain the experimental results is incorrect. Taking into account the above two effects, many other assumptions can explain the experiments. The results are as follows: TP0-0 (tensor+pseudoscalar and spin change 0-0) ST1-0 and VA1-0 fit the experiments only with the finite de Broglie wave length effect, A0-0 (including AP0-0) and VT1-0 requires both effects, but VT1-0 seems too artificial. Other cases can not fit the experiments.