Of course it's hard to make behavioral changes. That's part of the reason it's such a worthwhile goal.

It might be more effective if they were refactored in to a problem / solution format where it was easier for a reader to self-filter in to matching problem statements, then see a weighted list of correction methods that are more / less effective and try to implement a couple that might work for them.

Mind if I see some of this evidence? Because books really helped me, but only after I got myself into a much more accepting environment first, where I could experiment with the things in the books.