07 de desembre 2016

Motivated bayesians or classical bayesians

Classical Bayesians will both seek out the most informative evidenceand process it in an unbiased way. However, motivated bayesians gather and process information before and during the decision-making process and they tend to do so in a way that is predictably biased toward helping them to feel
that their behavior is moral, honest, or fair, while still pursuing their self-interest. This is the definition in an interesting article in JEP, and this is the summary:
First, we argue that people often form self-serving judgments of what, exactly, constitutes fair or moral behavior or outcomes. When there is some flexibility in interpreting what is “right” and “wrong” or “moral” or “immoral,” people’s judgments of the morality of an act are often biased in the direction of what best suits their interests. Second, we argue that a similar but distinct phenomenon occurs when people actually alter their judgments of objective qualities—such as their own abilities or the quality of competing options—as a way of making egoistic behavior appear more moral. Finally, we argue that motivated Bayesian reasoning in moral decision making has important implications for many behaviors relevant for economics and policy. In domains including
charitable giving, corruption and bribery, and discrimination in labor markets, the ability of people to pursue egoistic objectives while maintaining a belief in their own morality has important consequences for their behavior.
We argue that an underexplored element in much of this research is the frequent tendency of decision makers to engage in motivated information processing—acting as motivated Bayesians—thereby resolving the apparent tension between acting egoistically and acting morally. Individuals’ flexibility and creativity in how they acquire, attend to, and process information may allow them to reach the desirable conclusion that they can be both moral and egoistic at the same time.
The article is full of examples and you can add more evidence with this case nowadays in the press. At the end of the article, ask yourself if you are a motivated or classical bayesian, or maybe both according to context...

PS. Must read post on private and public health in India.