Making Descriptive Use of Prospect Theory to Improve the Prescriptive Use of Expected Utility Peter P. Wakker (& Bleichrodt & Pinto); Oct. 3, decision analysis publication award. This powerpoint file will be on my homepage on coming Monday. p? 1p1p Perf. Health artificial speech ~ EU = U = p Analysis is based on EU!?!? Standard gamble question: p 1 + (1–p) 0 = p Don’t forget to make this invisible. Already appeared two years ago. This conference already saw two follow-up studies. Hence, I speak about it for only 10 minutes. Rest of my time I use for a general discussion, about why I think that people have difficulties in using our corrections. I think, frankly, that people rather “look the other way” than face the biases.
2 Common justification: EU is normative (von Neumann-Morgenstern) We agree that EU is normative. But not that this would justify SG-analysis. SG measurement is descriptive. EU is not descriptive. A better descriptive theory: Prospect theory!
3 EU: U(x) = p. PT: U(x) = p p + (1 p) w + ( ) ww We: is wrong !!
p w+w /3 Figure. The common weighting fuction w + (1/3) = 1/3; 4 = 2.25 w is similar;
Standard Gamble Utilities, Corrected through Prospect Theory, for p =.00,..., E.g., if p =.15 then U = 0.123
U p Corrected Standard Gamble Utility Curve 6
U SG U CE ( at 1 st = CE(.10), …, at 5 th = CE(.90) ) 5 th 3d3d 1 st 2 nd 4 th *** * ** * *** * Corrected (Prospect theory) U SG U TO ( at 1 st = x 1, …, at 5 th = x 5 ) U CE U TO ( at 1 st = x 1, …, 5 th = x 5 ) Classical (EU) 7
SG(EU) CE 1/3 SG(PT) SP TO Utility functions (for mean values). 0 1/6 2/6 3/6 4/6 5/6 1 7/6 t 0 = FF5,000 FF U 8 t 6 = FF26,068 Abdellaoui, Barrios, & Wakker (2003)