"Science of the psyche" is basically experimental psychology and behavioral science. All these use statistics to analyze data, as well as basic neuroscience. Of course anyone versed in statistics can choose or design a model/test that will more likely to support their pet hypotheses. Well, almost. Statistics is based on math, but the models are highly subjective and can be manipulated easily.

I think the OP asked about science showing that psychotherapy works, so not only the negative outcomes.