Question: Ellen, a risk-neutral test taker, just finished taking her decision analysis midterm. She understood the scoring rule from Homework #5, but was perplexed when she received a - 2c on the exam. On one of the questions she was absolutely certain that the answer was "b," so she assigned her belief of p = s = 1 to "b." Unfortunately for Ellen, she marked the wrong box and inadvertently put one for "c" and a zero for everything else, and as a result, p = 0. Oops. Now, Ellen wants you to help her identify how she should approach next year's midterm

Suppose that during the exam, she believes that {b correct|&} = 5 and {a correct|&} = {c correct|&} = {d correct|&} = (1 - .v)/3. She would like to add another distinction, named "oops, I made a mistake" with two degrees "ouch that hurt" and "whew, I'm ok after all," which we'll just denote M and M', respectively. Let {M | b correct,& } = q. This all can be represented as the following deal.

For simplicity, assume that she equally distributes the probability weights among the other choices (a, c, and d.)

Suppose Ellen is a deltaperson with risk odds r = 1/2. How do your results change? Do they make sense given this attitude toward risk?