Tuesday, January 24, 2012

Limits of Heuristics?

In his elegant book, Thinking Fast and Slow (Farrar, Straus, and Giroux 2011), Daniel Kahneman reports on many fascinating findings of cognitive science over the last half century, including the following: when confronted with difficult questions, which would normally require a "System 2," deliberative approach requiring relatively greater attention and energy, individuals often resort instead to answering substitute, "heuristic" questions that allow System 1, the fast, intuitive decision process, to come up with quick and dirty solutions. Here is one of his examples: when faced with the "target" question, "How popular will the president be six months from now?" individuals will substitute the "heuristic" question, "How popular is the president now?" Because System 1 is, in fact, answering a different question, there is a heightened risk that the answer will not be correct for the more difficult target question. Nevertheless, in many cases the "heuristic alternative to careful reasoning ... works fairly well" (p. 98). While System 1's heuristic answers are subject to review and rejection by System 2, "a lazy System 2 often follows the path of least effort and endorses a heuristic answer without much scrutiny of whether it is truly appropriate."

I do not doubt that Kahneman's description of this process is correct (I certainly have no basis for raising any such doubts). What I wonder, however, is whether individuals typically differentiate between low-stakes and high-stakes questions/decisions, calling on System 2 when the stakes or risks stemming from an erroneous answer are relatively high. Kahneman doesn't raise this question (at least not at this point in the book). I suspect we do call on System 2 more as the costs of erroneous answers rise. If so, then we require some basis for explaining how individuals differentiate high-stakes from low-stakes circumstances, which the brain must be able to do very, very quickly.

I have not read past Chapter 9 in the book yet, and perhaps Kahneman takes up my question in later pages (although I didn't find it on a quick leaf through the balance of the book). However, given his findings about the generally poor ability of individuals to calculate risk, one wonders how individuals could accurately decide when the stakes are high enough to call on System 2.

I don't know the cognitive psychology literature at all well - certainly not well enough to know whether anyone's tested experimentally whether raising the costs of incorrect answers, or rewards for correct answers, to "target questions" reduces reliance on System 1's intuitive responses to substitute, heuristic questions. I would be surprised if no one has done this. I would be even more surprised to learn that material incentives play no role at all in the way individuals make decisions.

No comments:

Post a Comment

I actively moderate comments for spam, advertisements, and abusive or offensive language.