I'm rereading Dr. Clay's book, looking for things that jump out at me at this time

In Chapter 1 he was talking about self help research and recent research and I wondered about that. So I went looking and found this:

http://www.selfhelpweb.org/research.html

Has any one thought about how one would research "self" help? It's not like you can force another to try to help their self or even want to help their self and if one is doing self help, it's too late? What things work for one person might not work for another. So, if a self help method only helps 2% of those who seriously try it, do you say it isn't any good? Those 2% would be mighty unhappy if it hadn't been suggested, offered or thought of wouldn't they?

But what IS research into self help? Does one research whether it is helpful? or how one chooses a particular method or idea? or what sorts of problems one is having that one chooses that method? Are there some things that self help won't work for?

Dr. Clay starts from the position that most behaviors are learned and because they are learned they can be unlearned or replaced with other learning or added to to reinforce, etc. I added the last bits there; what is the nature of "help"? Is there a difference between quitting smoking, say, and quitting SI? Not biting one's fingernails and not cussing? Going to church each week and going to psychotherapy? Help stopping something one perceives as negative verus help starting and doing something one perceives as positive? Why "help"? Is "just do it" self help?

Why would a researcher research self help?