учебник
задачник
вход
РУС
ENG

Институты

 кейсы

 эссе

 задачи

 закрытые вопросы

 

Институты
кейсы

Проверка гипотезы рациональности: эффект постановки вопроса при выборе способа лечения

 Эксперимент 1.

«[This] example comes from a study of preferences between medical treatments. Respondents were given statistical information about the outcomes of two treatments of lung cancer. The same statistics were presented to some respondents in terms of mortality rates and to others in terms of survival rates. The respondents then indicated their preferred treatment. The information was presented as follows.

Problem 1 (Survival frame)

Surgery: Of 100 people having surgery 90 live through the post-operative period, 68 are alive at the end of the first year and 34 are alive at the end of five years.

Radiation therapy: Of 100 people having radiation therapy all live through the treatment, 77 are alive at the end of one year and 22 are alive at the end of five years.

Problem 1 (Mortality frame)

Surgery: Of 100 people having surgery 10 die during surgery or the post-operative period, 32 die by the end of the first year and 66 die by the end of five years.

Radiation therapy: Of 100 people having radiation therapy none die during treatment, 23 die by the end of one year and 78 die by the end of five years.

The inconsequential difference in formulation produced a marked effect. The overall percentage of respondents who favored radiation therapy rose from 18% in the survival frame (N = 247) to 44% in the mortality frame (N = 336). The advantage of radiation therapy over surgery evidently looms larger when stated as a reduction of the risk of immediate death from 10% to 0% rather than as an increase from 90% to 100% in the rate of survival. The framing effect was not smaller for experienced physicians or for statistically sophisticated business students than for a group of clinic patients».

(The number of respondents who answered each problem is denoted by N)

Эксперимент 2.

«Problem 5 (N = 152).

Imagine that the U.S. is preparing for the outbreak of an unusual Asian disease, which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed. Assume that the exact scientific estimates of the consequences of the programs are as follows:

If Program A is adopted, 200 people will be saved. [72%]

If Program B is adopted. There is 1/3 probability that 600 people will be saved, and 2/3 probability that no people will be saved. [28%]

Problem 6 (N = 155):

If Program C is adopted 400 people will die. [22%]

If program D is adopted there is 1/3 probability that nobody will die, and 2/3 probability that 600 people will die [78%]»[1].(The number of respondents who answered each problem is denoted by N, and the percentage who choose each option is given in brackets.)


Вопрос

Какой бы выбор сделал совершенно рациональный агент? Почему в ходе эксперимента выбор людей отличался от выбора совершенно рационального агента?

Какие предпосылки ограниченной рациональности позволяют сделать вывод об ограниченной рациональности участников эксперимента?


Ответ

 Kahneman and Tversky explain:

«The respondents … exhibited the standard pattern of risk aversion in gains and risk seeking in losses. … As illustrated by the preceding examples, variations in the framing of decision problems produce systematic violations of invariance and dominance».

«In problem 5 the outcomes are stated in positive terms (lives saved), and the majority choice is accordingly risk averse. The prospect of certainly saving 200 lives is more attractive than a risky prospect of equal expected value.

A second group of respondents was given the same cover story with the following descriptions of the alternative programs.

In problem 6 the outcomes are stated in negative terms (lives lost), and the majority choice is accordingly risk seeking. The certain death of 400 people is less acceptable than a two-thirds chance that 600 people will die. Problems 5 and 6, however, are essentially identical. They differ only in that the former is framed in terms of the number of lives saved (relative to an expected loss of 600 lives if no action is taken), whereas the latter is framed in terms of the number of lives lost»


Сноски

 Tversky A., Kahneman D. Rational Choice and the Framing of Decisions // Journal of Business. 1986. Vol. 59. № 4. Part 2. P. 255.

Ibid. 260

Tversky A., Kahneman D. Rational Choice and the Framing of Decisions // Journal of Business. 1986. Vol. 59. № 4. Part 2. P. 255.


О проекте
Руководство пользователя
Авторский коллектив
Контакты
Задачник
Учебник
Видео
Форум
Бумажный учебник
Мы в СМИ
Загрузить Econline
Лицензионное соглашение
Технологии