8 Eylül 2009 Salı

Expansion

Maybe you remember, a while ago I wrote a book comment on my blog. The book called Improbable... I have quoted a part from the book where Pascal's decision about being religious or not being religious is done by a probabilistic method called expected value. Honestly, I did not know that this expected value analysis which is so common for statistics and probability is not enough to explain some real world applications. But actually I thought that if it was so obvious, not only Pascal but so many people could have made similar decision. Now I know why they can not do and I figured it about by just a coincidental reading.
...
Kahneman and Tversky who got Nobel prize in economics found out the violations of the expected value theory.
1) Individuals are often observed overweighting outcomes considered certain relative to those which are probable. It means that weighting the probabilities is not due to objectivity but rather due to their sense of likelihood. It was found out that people have hard times to realize the weights of the events which are almost certain and almost impossible. For them, they either certain or impossible. On the other hand, people overweight the low probability events and underweight the medium and high probability ones.
2) While individuals exhibit risk averse behavior over potential gains, they show risk seeking behavior over potential losses.
3) The disutility of a loss is larger in absolute terms than the utility of a gain of equal magnitude. Which means that loss is affecting the behavior more than gain even though they have the same absolute values. Being loss averse...
4) When faced with a pair of prospects, subjects seem to decompose them into their common and different components, and base their choice on the components that are different, thus “isolating” the common components as irrelevant for decision.
5) Excessive discounting of the distant future relative to the near future. Myopia...
...
All in all, it means that people have 'bounded rationality' in the sense of violating the axioms of expected utility which causes them not being able to find efficient solutions for decisions.

Hiç yorum yok: