I recommend this course because it is a fun introduction covering many of the examples that you must work your way through in order to understand how much of what we take as immediate and direct requires learning and processing.  What if that positive manifold, the halo effect, was not measurement error but a real manifestation of a greedy algorithm supporting action in an uncertain environment?  If such were the case, then I would not want my statistical model to control for halo effects because the halo is the driver of behavior.  Similarly, I would not use self-reports in my statistical model if the self-reports were after-the-fact justifications of choices made for other reasons.  These are the issues raised by Dan Ariely’s Coursera course.  They ought to be addressed in our model specification.