Here is an unpleasant fact about ’real life’ data. OWC (or GARP) is almost always violated. In any actual data set, be it from the laboratory or from the ’real world’, individuals will almost certainly fail the relevant axiom. Remember, one mistaken choice, one slip up, and the whole data set will fail OWC. This is problematic, as this is not a very interesting result: if we are going to classify everyone as irrational, then do we throw out all the machinery of economics, possibly due to a very small number of rogue choices? This seems too strong. Therefore, it would be nice to have some measure of how close a particular data set is from satisfying rationality. In this section, we are going to present some tools that try and put a metric on these things. It should be noted that these metrics are somewhat arbitrary, and in general lack a proper statistical grounding. A fruitful area of research might be to put these things on a firmer footing.1
分享
引用
