Check out this insightful (imo) take on how behavioral economics can address normative questions. It argues against assumptions of people as hyper-rationalistic agents as bases for economic models.
I suspect this tendency is true at a layman level as well. On one hand (normal) people understand and rationalize what is considered (- and all too common in conversations -) “human nature”, but when posed normative questions, people apply rationalistic principles. I think this also contains a pointer to how people’s responses to hypothetical scenarios from questionnaires do not necessarily represent their behaviour when actually faced with said situations. Example: A lack of sales of goods for which people indicated demand in market surveys.
(Also, this paper made me speculate a bit on how hypnosis works.)