Andrew Gelman mentioned “the Kahneman-Gigerenzer catfight, or more generally the endless debate between those who emphasize irrationality in human decision making and those who emphasize the adaptive and functional qualities of our shortcuts.” This looked worth checking, so I followed the link to the following statement by Gigerenzer:
The “half-empty” versus “half-full” explanation of the differences between Kahneman and us misses the essential point: the difference is about the nature of the glass of rationality, not the level of the water. For Kahneman, rationality is logical rationality, defined as some content-free law of logic or probability; for us, it is ecological rationality, loosely speaking, the match between a heuristic and its environment. For ecological rationality, taking into account contextual cues (the environment) is the very essence of rationality, for Kahneman it is a deviation from a logical norm and thus, a deviation from rationality. In Kahneman’s philosophy, simple heuristics could never predict better than rational models; in our research we have shown systematic less-is-more effects.
LW’s dog in this catfight is probably on the Kahneman’s side, but the debate is interesting.
LW’s dog in this catfight is probably on the Kahneman’s side
Well, probability is about reasoning with logic under imperfect information, and when you factor in the cost of elaboration you see that “ecological” model could be better, but evolution and thermodynamics. I think that simply distinguishing “correct” and “useful” dissolves the debate.
I think that simply distinguishing “correct” and “useful” dissolves the debate.
No, I think it’s more complicated than that.
For example, imagine a complex decision, say what college to go to. Can you write out a Bayesian model that will tell you what to do? Well, kinda. You can, but it’s going to be woefully incomplete and involve a lot of guesses without much support from data. A set of heuristics will do much better in this situation. Are you going to say that this Bayesian model is “correct” regardless? I don’t think it’s a useful application of the word.
Not necessarily. “You can’t do inference without making assumptions”.
Is it even a fight? What is it that they disagree about? Neither side is saying “Decision heuristics that once worked well still work well in our changed world”.
Eh. This is sounding more and more like a dispute over definitions, and hence tedious; and I would be unsurprised to find that it arose from either self-promotion or ideology; q.v. the Gould and Eldredge v. Maynard Smith et al. kerfuffle.
The two approaches might but not necessarily will lead you to the same thing. I suspect that part of the tension is between “theoretically correct” and “works better in practice” which in theory should match but in practice do not often enough.
Here is what looks to be the major Gigerenzer paper.
Something tells me Gigerenzer is misquoting Kahneman, he is just saying any deviation from that counts as irrational and measuring that as his baseline, i’m more than sure he would be happy to use ecological rationality as a baseline as well.
Andrew Gelman mentioned “the Kahneman-Gigerenzer catfight, or more generally the endless debate between those who emphasize irrationality in human decision making and those who emphasize the adaptive and functional qualities of our shortcuts.” This looked worth checking, so I followed the link to the following statement by Gigerenzer:
LW’s dog in this catfight is probably on the Kahneman’s side, but the debate is interesting.
Well, probability is about reasoning with logic under imperfect information, and when you factor in the cost of elaboration you see that “ecological” model could be better, but evolution and thermodynamics. I think that simply distinguishing “correct” and “useful” dissolves the debate.
No, I think it’s more complicated than that.
For example, imagine a complex decision, say what college to go to. Can you write out a Bayesian model that will tell you what to do? Well, kinda. You can, but it’s going to be woefully incomplete and involve a lot of guesses without much support from data. A set of heuristics will do much better in this situation. Are you going to say that this Bayesian model is “correct” regardless? I don’t think it’s a useful application of the word.
Not necessarily. “You can’t do inference without making assumptions”.
Is it even a fight? What is it that they disagree about? Neither side is saying “Decision heuristics that once worked well still work well in our changed world”.
It’s a fight like a croquet mallet is a billy club :-)
I mean, is there some prediction that they disagree about, rather than ‘falling tree sound’ issues.
Eh. This is sounding more and more like a dispute over definitions, and hence tedious; and I would be unsurprised to find that it arose from either self-promotion or ideology; q.v. the Gould and Eldredge v. Maynard Smith et al. kerfuffle.
I admit that I don’t get the explanation. Wouldn’t both approaches lead to the same thing?
The two approaches might but not necessarily will lead you to the same thing. I suspect that part of the tension is between “theoretically correct” and “works better in practice” which in theory should match but in practice do not often enough.
Here is what looks to be the major Gigerenzer paper.
Something tells me Gigerenzer is misquoting Kahneman, he is just saying any deviation from that counts as irrational and measuring that as his baseline, i’m more than sure he would be happy to use ecological rationality as a baseline as well.