1、Section 1 The Psychology of PreferenceScientific American 264，PP 160173by Daniel Kahneman and Amos Tverskycontent 1.1 Overview 1.2 Risk aversion and Risk-seeking 1.3 Value function 1.4 Decision weights vs probability 1.5 Framing effect 1.6 Reference point 1.7 Regret 1.8 Conclusion1.1 Overview When p
2、eople make risky choices,they do not do so objectively.Experiment:Broadway play lost the ticket(paid$40):buy another one?lost$40:Will you buy ticket?Interpretation:different mental account Discrepancies between subjective and objective conceptions of decisions Examples:threat of loss vs equivalent g
3、ain Sensitvie:Difference between certainly and high probability Insensitive:Difference between intermediate gradations of probability Regret with a loss by action and that of inaction 1.2 risk aversion and risk-seeking The origin of psychology of preference Daniel Bernoulli(1738)discussed the charac
4、teristics of human preferences:risk aversion Example:option 1:a sure gain of$80 Option 2:risky prospect 85%winning$100,15%winning nothing A choice is risk-averse if a certain outcome is preferred to a gamble with an equal or greater monetary expectation Example:Option 1:a sure loss of$80 Option 2:a
5、risky outcome 85%losing$100,15%losing nothing A choice is risk-seeking if a certain outcome is rejected in favor of a gamble with an equal or lower monetary expectation Conclusion Preference between gains are risk-averse Preference between losses are risk-seeking People forgo the option that offers
6、the highest monetary expectation.WHY?Bernoullis answer:expected utility utility is not a linear function of money1.3 Value function How do people identify the outcomes of a decision?States of wealth vs changes of wealth Comprehensive view of vs limited view Identify consequences as gains or losses r
7、elative to a neutral point,which leads to inconsistent choices(because the same objective consequence can be evaluated in more than one way)Definition of value function:a function that associates a subjective value to any objective amount that may be gained or lost.Concave value function(gains)Assum
8、ption:Gains have positive value Zero gains has a subjective value of zero The value function for gains are concave downward(each extra dollar gained adds less to value than the preceding one)Shallow region of the function Steepest region of the function Value is nonlinear,the sure gain is closer to
9、the large gain in terms of value than in terms of moneyConvex value function(losses)Assumptions:Losses have negative value The value function for losses is convex(each extra dollar loss causes a smaller change in value than the preceding one)Flatter region of the function Steepest region of the func
10、tion The sure loss is relatively closer to the worst outcome on the value scale than it is on the money scaleMathematical estimation Example:A sure cash prize 50%winning$100+50%winning nothing What amount would make the sure prize just as attractive as the bet?Value$35 vs half the value of$100 50%wi
11、nning$200,$500,$1000,$2000 The prize is roughly proportional to the stake Example 50%losing$100 +50%losing nothing What amount would make the sure loss just as acceptable as the bet?Value-$40 is half the value of-$100 Power function is a good estimates-shaped value function(gain&loss)Example 50%losi
12、ng$100 50%winning a cash prize What is the smallest prize that would make this bet acceptable?Pleasure of winning a sum of money is much less intense than the pain of losing the same sum The asymmetry in the response to gains and losses is expressed in the greater steepness of the value function for
13、 losses.Individuals differ in their attitudes toward risk and toward money,therefore no single value function can describe the preferences of all individuals The near proportionality of prizes to stakes breaks down beyond the range of moderate gains and losses.1.4 decision weights vs probability Exa
14、mple Imagine that you can improve your chance to win a very desirable prize.Would you pay as much to raise your chance from 30%to 40%as you would to raise your chance from 90%to certainty?The difference between certainty and possibility and the difference between possibility and impossibility loom l
15、arger than comparable differences in the intermediate range of probability Overweighted (low probability)Underweighted(intermediate and high probability)Impossibility weighted 0 Certainty weighted 1 The overweighting can lead to risk seeking in the positive domain Lottery and accident insurance The
16、underweighting can lead to risk aversion in the negative domain If that is the case,the objective probability will be replaced by a slightly smaller decision weight.The value function would become less curved,still S shaped1.5 Framing effect How people define the consequences of their choice?The sam
17、e decision can be framed in several different ways,Different frames can lead to different decisions Example You have been given$200 You are asked to choose A.a sure gain of$50 B.25%winning$200+75%winning nothing A is prefered (risk averse)Example You have been given$400 You are asked to choose C.a s
18、ure loss of$150 D.75%losing$200 +25%losing nothing D is preferred (risk seeking)In fact,the options in the above two problems is identical in objective terms The choice of the gamble in either winning yields a 75%of winning$200 and 25%of winning$400.Rational decision:combine the bonus with the avail
19、able options.In fact,they ignore the bonus and evaluate the first one between gains and the second one between losses The reversal of preference is induced by altering the description of outcomes.We call such reversals framing effects.1.6 reference point Framing effects arise when the same objective
20、 alternatives are evaluated in relation to different points of reference Example Two alternative programs A.200 will be saved B.1/3 600 will be saved+2/3 no one saved A is preferred(risk averse)Example C.400 die D.1/3 no one dies+2/3 600 die D is preferred(risk seeking)In fact,the above two versions
21、 is identical The first one takes death of 600 as RP and takes the lives saved as gains The second one takes no death as RP and takes the lives lost as loss Because of the S-shaped value function and the overweighting of certainty,the frames yield different preferences Single dimension of value Seve
22、ral dimension of value:transaction A transaction must be evaluated according to the balance of cost and benefits in a mental account.Example Jacket for$125 Calculator for$15 Calculator on sale at other store for$10(20 minutes drive away)Would you make a trip to the other store?Answer:Yes Example Jac
23、ket for$15 Calculator for$125 Calculator on sale at other store for$120(20 minutes drive away)Would you make a trip to the other store?Answer:No In fact,they the same:One has to decide whether to drive 20 minutes to save$5 The first example:people compare the$5 with the price of calculator($15),whic
24、h is more impressive than a reduction from$125 to$120 Framing effect may be particularly pronounced in situations that have a singe dimension of cost(money)and several dimension of benefit.Example:Tape deck vs a new car Buyers of homes vs furniture1.7 regret What determines the reference point?Somet
25、imes:The RP is the state to which one has become adapted IN many cases,the RP is determined by events that are only imagined.Example:Mr.Crane Mr.Thomas Who is more upset?(Thomas)In fact,both missed the flights Example:winning number 865304 John 361204 Mary 965304 Peter 865305 Who is more upset?(Pete
26、r)Example Paul stock of company A George from stock B to stock A Who is more upset?(George)The anticipation of regret is likely to favor inaction over action and routine behavior over innovative behavior1.8 conclusions Some of the above examples illustrate the impediments to the rational decision,that were discussed by A.Simon under the heading of“bounded rationality”.
