I don't know whether the Amos model you use is a latent variable model or a path model. As far as I know, the principle of path model is the same as that of SPSS, and the results are very close, but the latent variable model is not necessarily the case. If the paper has explained the use of Amos, then it can't be said that the second best thing is to use SPSS, because Amos, as a more comprehensive and rigorous method, has rejected the research hypothesis, and it is obviously meaningless to test the hypothesis with an imprecise method.