Скачать презентацию Decision making Definition Decision making two or

2dd5386aaeae4f69fd05c6a9aa246d4c.ppt

• Количество слайдов: 50

Decision making

Definition Decision making: two or more choices laid out, and you must select one QUESTION: what, if any, are the consistent rules that guide such decisions?

Decision making You would think that people want to get the most goods, \$\$, etc. , and to minimize bad things. That’s true as far as it goes, but what about when the decision gets complicated. . . In that case, you might think that people are “conditioned” by experience to make the optimal decision.

Summary of main point: • People’s reasoning is not driven by optimality. • People’s reasoning is driven by heuristics.

Rational theory: refers to being internally consistent. There are definitely situations in which people do not behave rationally (e. g. violate transitivity). A>B, B>C, therefore A>C

Normative theory: theory is prescriptive, values some choices over others. E. g. , expected value theory--get the highest expected value. (Probability obtain) X (value) E. g. , . 5 chance of getting 25\$: E. V. = \$12. 5 E. g. , . 25 chance of getting 55\$: E. V. = \$13. 75

So does this mean that people always choose the highest expected value?

Calculating value can be hard: how do you choose an apartment?

Rank them How close? Rent Laundry Room size Apt 1 1 2 3 2 8 Apt 2 3 4 2 3 12 Apt 3 4 1 1 1 7 Apt 4 2 3 4 4 13 Stupid—not all factors are equally important

Rank them How Rent (. 6) close? (. 8) Laundry (. 2) Room size (. 2) Apt 1 1 2 3 2 8 3. 0 Apt 2 3 4 2 3 12 5. 8 Apt 3 4 1 1 1 7 4. 2 Apt 4 2 3 4 4 13 5. 0

More formal violation of expected value

. 80 chance of winning 25\$: E. V. = \$20. 40 chance of winning 55\$: E. V. = \$22 Expected utility theory--value of outcomes vary depending on the individual (Probability win) X (value of prize) You might prefer the lower e. v. , if you had some special need for 25 dollars, for example. The utility of the extra 30 is low.

e. v. =\$1. 00 A) 1. 0 chance of winning \$1 B) B). 00000014 chance of winning \$3 million: e. v. = \$0. 42 The utility of 3 million outweighs the disadvantage in expected value.

Irrationality: eliciting preference (a) 8 chances in 9 to win \$4 (b) 1 chance in 9 to win \$40 Method 1: which would you pick? 71% pick (a) Method 2: What’s the lowest price at which you would sell the chance to play this game? 67% assign a higher price to (b)

Eliciting preference #2 You are on the jury of an only-child sole-custody case following a messy divorce. You decide to base your decision on just a few observations. . . Award Deny Parent A Average Income Average health 36% 45% Average work habits Reasonable rapport with child Relatively stable social life Parent B Above average Income Minor health problems 64% Lots of work related travel Very close relationship with child Extremely active social life 55%

Examples of Problem Changes • Problem frame effects • Psychic budgets • Sunk costs

Irrationality: framing Do you prefer: (a) 50% chance of winning \$200 and 50% chance of winning \$50 (b) receiving \$100 for sure Risk averse for gains (a) 50% chance of losing \$200 and 50% chance of winning \$0 (b) losing \$100 for sure Risk seeking for losses

Problem frame A) 200 people will be saved. B) There’s a 1/3 chance that 600 people will be saved, and a 2/3 chance that no people will be saved A) 400 people will die B) There’s a 1/3 chance that no one will die, and a 2/3 chance that 600 people will die.

(72%) A) 200 people will be saved. (28%) B) There’s a 1/3 chance that 600 people will be saved, and a 2/3 chance that no people will be saved (22%) A) 400 people will die (78%) B) There’s a 1/3 chance that no one will die, and a 2/3 chance that 600 people will die.

Psychic Budgets Psychic budgets concern how we mentally categorize money we have spent or are contemplating spending. The categories don’t always correspond to reality, and so can lead to odd decisions.

Psychic Budgets Suppose you and a friend are going to theater. When you get to theater you realize that somewhere on your way you’ve lost the tickets, which cost a total of \$100. You have more than \$100 cash with you. Tickets are still available at the box office. Would you buy tickets to replace those that you’ve lost? Suppose that you and a friend are going to theater. When you get to theater you realize that somewhere on our way you’ve lost \$100 cash. The tickets cost \$100, and you can get a full refund. Would you cash in the tickets to offset the loss?

Sunk costs A sunk cost is time, money, or other investment that is irretrievably spent, and therefore should not affect current decision making, and yet does.

In order to “get your money’s worth” you: finish a crummy meal. In order to “get your money’s worth” you sit through a stupid movie.

Summary The foregoing material shows us that people don’t always behave rationally. That fact implies that some principle other than expected utility guides their choices. What is that principle?

Heuristics A heuristic is a simple cognitive rule that is easy to apply and usually leads to acceptable answers, but can also lead to errors. • • Representativeness Availability Anchoring and adjustment Simulation

Representativeness is used when you are asked to judge the probability of an event. Representativeness leads you to judge that an event is likely to belong to a category if the event either has the features of the category that you strongly associate with that category, or if it has features that you associate with the process that generates examples of the category

Representativeness Which coin toss is more likely? Many people pick the bottom one, because it looks more representative of randomness, the process that people know governs coin tosses.

Another example Linda is 31 years old, single, outspoken and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in antinuclear demonstrations. Which of these is more likely? Linda is a bank teller and is active in the feminist movement

Availability is used to judge the probability of events You use availability by trying to call to mind examples: the more examples you think of, the more probable you judge the event to be.

Availability Classic example: are there more words with “R” as the first letter or with “R” as the third letter?

Availability More likely cause of death: Airplane crash Auto crash Diabetes Homicide

Estimates of causes of death • • • Childbirth/abortion Tornadoes Flood Botulism Fire Homicide • Venemous bite or sting • • Smallpox vaccination Diabetes Stomach cancer Stroke Tuberculosis Asthma Emphysema

Implication Portrayal of racial and ethnic groups on television /movies; e. g. , African Americans over-represented as criminals on TV; availability heuristic suggests you’ll believe. . .

Anchoring and adjustment is used to estimate probabilities To use it, you do a quick estimate based on memory, or on information provided in the problem, and then adjust the estimate.

Anchoring and adjustment Classic example: you have 5 seconds to answer this problem: 1 X 2 X 3 X 4 X 5 X 6 X 7 X 8 (median guess = 512) OR 8 X 7 X 6 X 5 X 4 X 3 X 2 X 1 (median guess = 2, 250) People multiple the first few numbers (anchor) and then adjust that number. But adjustments are usually inadequate.

Anchoring and adjustment “You’re not going to pay \$130 for this food dehydrator. ” Pitchman is providing you with an anchor. You adjust downwards, but adjustments are usually inadequate, so when you see the final price, it seems low.

Simulation Mental construction or imagining of outcomes--used to evaluate past outcomes, which may influence future decisions.

Simulation Downhill--adding a high-probability event (that didn’t happen) to the mental simulation so that the outcome would have been different. Horizontal--trading equal probability events so as to change the outcome Uphill--adding low probability events

Simulation Two guys arrive at the airport 30 minutes after their planes were supposed to leave. One finds out his plane was delayed. He just missed it by five minutes. The other guy’s flight left on time. Who is madder?

Simulation Application--slot machine

Information that we ignore Just as important to our decision-making processes are types of information that we systematically ignore. • Sample size • Base rate

Sample size refers to the number of things in a group that we are evaluating. Sample size matters to the accuracy of an estimate, but people sometimes don’t take account of that. For example. .

Sample size A certain town is served by two hospitals. In the larger hospital about 45 babies are born each day, and in the small hospital about 15 babies are born each day. As you know about 50 percent of all babies are boys. However, the exact percentage varies from day to day. Sometimes it may be higher than 50 percent, sometimes lower. For a period of 1 year, each hospital recorded the days on which more than 60 percent of the babies born were boys. Which hospital do you think recorded more such days? The larger hospital [21] The small hospital [21] About the same (that is, within 5 percent of each other) [53]

Sample size Other times, people are sensitive to sample size: • A new bird, shreeble is blue. . . • A new islander is obese. . .

Why do people use sample size sometimes and not others? People appear to be more likely to use sample size when the role of chance/consistency is highlighted in the problem, or in their experience.

Base Rate I have a friend who is a professor. He likes to write poetry, he’s rather shy and is small in stature. His hobby is raising rare orchids. Which field do you think he’s in? (a) Chinese studies (b) Psychology

Base Rate In a certain city there are two cab companies, the Blue and the Green. In this city the Blue company owns 85% of the cabs and the Green owns 15%. A cab is involved in a hitand-run accident. An eyewitness says she thinks it was a Green cab. The eyewitness’ vision is carefully tested and it is determined that, under the lighting conditions at the time of the accident, she can correctly identify the color of the cab 80% of the time. What are the odds that the hit-and-run cab was green?

Base rate Many people say 80% green--the presumed accuracy of the eyewitness. What would you say, though, had there not been an eyewitness? You’d say 85% blue, because of the base rate. This base rate information does not become irrelevant once there is added information from the eyewitness

Base rate Once given individuating information, people are more likely to ignore base rates.

Summary • We don’t reason based on logic • We reason via the application of heuristics • There are important sources of information that we systematically ignore when we reason.