Since you think that the research that investigates the effectiveness of positive thinking (such as this one: http://www.psych.nyu.edu/oettingen/Barry%20Kappes,

Since you think that the research that investigates the effectiveness of positive thinking (such as this one: http://www.psych.nyu.edu/oettingen/Barry%20Kappes,%20H.,%20&%20Oettingen,%20G.%20%282011%29.%20JESP.pdf http://www.psych.nyu.edu/oettingen/Barry%20Kappes,%20H.,%20&..., linked from the article) has a methodological defect, how would you design an experiment so that we could correct for the problem? That is, your point is that the research conducted over-simplified the technique. If I understand your point, "thinking positively" is not enough; the subjects in the experiment did one step, but not the others that you think are required. So how do you design an experiment where thinking positively necessarily motivates "one to build successful habits"?