How to use the tradeoff table:

Imagine that your physician tells you that have one of two possible diseases.  You have an equal chance of having either Disease 1 or Disease 2. However, your physician is unsure as to which disease you actually have.

Despite the uncertainty as to your condition, your physician thinks it is necessary to proceed with some form of treatment. There are two treatments that can help you.  The table below shows the years of life you will obtain under each disease if you choose Treatment 1 or Treatment 2. The green cell represents the number of years of life Treatment 2 will give you if you have Disease 2. This cell is blank and requires of you the following judgment.

Please enter in *green* cell of the table, the number of years of life that Treatment 2 would have to give you if you were to have Disease 2, in order for the two treatments to be of equal preference.  Then press <Continue> button.  When you press this button, some of the possible outcomes will change.  Please repeat the exercise until all values of the years column have been determined..

Copyright © 2001 by Jason N. Doctor.   All rights reserved . This program may be used freely only for educational and other noncommercial scholarly uses. Software is “as is,” no guarantees or warranties can be made. You may modify the program, but retain this copyright notice.

About the UTILs scale:  The UTILs scale is an interval scale of utility that is derived from utility differences.  The scale is set such that U(0yrs) = 0 utils.  Each subsequent unit of life years utility is determined by the subjectys response to the tradeoff question.  From the TO exercise one gets a rough estimate of the shape of your utility function over life years.  An advantage of the TO method is that it does not require the explicit use of probabilities.  Another advantage is that it does not require assumptions about the parametric form of the utility function (or probability weighting function).  A disadvantage is that it is susceptible to accumulation of error.

Reference:  Wakker, P.P. and Deneffe, D. 1996.  Eliciting von Neumann-Morgenstern utilities when probabilities are distorted or unknown.  Management Science, 42, 1131-1150.