As is my usual habit, I assumed that you could manage elementary reading comprehension. Now, are you going to answer my point?
Wrong on all points, actually. here is one that isn't paywalled.
An article that I was first to link in this discussion. The data clearly show that recidivism is mainly predicted by the attributes of the prisoner (Annex A) while interventions have little effect (Annex F).
Note the decidedly non-zero effect. And note also that prisoner attitudes can be, and are, changed.