You can start by checking out research on subliminal messaging. To summarize, subliminal messaging works to some degree, but it is much less effective than superliminal communication.
For example, will showing a dark gray coke bottle silhouette, so dark that you don't even know you're seeing it, in the momentary blank screen between movie previews cause people to decide to go buy a soda? It will probably influence a couple people, but not as many as would be influenced by just replacing the blank screen with a big, bright, picture of a coke, with the Coca-cola logo and condensation dripping down the glass. I say this because the benefits of conventional advertisements are well-established, but the effects of subliminal marketing are in doubt (Karremans et al., 2006).
That said, there is evidence that you can "prime" someone with subliminal information. Priming, though, is more about getting people to recognize something more quickly later than it is about getting them to change their behavior.
To your question, Greenwald has studied the effectiveness of subliminal messaging for self-help, and found it to be ineffective.
Refs:
Greenwald, A. G. (1992). New Look 3: Unconscious cognition reclaimed. American Psychologist, 47(6), 766.
Karremans, J. C., Stroebe, W., & Claus, J. (2006). Beyond Vicary’s fantasies: The impact of subliminal priming and brand choice. Journal of Experimental Social Psychology, 42(6), 792-798.
