Judgment & Decision-Making Chapter 13 PDF

Judgment & Decision-Making Chapter 13  Overview 1. What is decision-making? 2. How do people make decisions? 2.1. Normative approaches 2.2. Descriptive approaches 2.3. Prescriptive (heuristic) approaches 2.3.1. Representative heuristic 2.3.2. Availab...

Judgment & Decision-Making Chapter 13  Overview 1. What is decision-making? 2. How do people make decisions? 2.1. Normative approaches 2.2. Descriptive approaches 2.3. Prescriptive (heuristic) approaches 2.3.1. Representative heuristic 2.3.2. Availability heuristic 2.3.3. Anchoring & adjustment 3. Fun but advanced: Nudges, Decision Architectures, Game Theory, Emotions, Intertemporal Choice  1. What is decision-making? Choosing amongst alternatives Properties of a decision: – Mutually exclusive alternatives – Future consequences / different courses of action – Different values It’s all around us:  2.1. Normative approaches Expected value (EV): Throw a die. If it comes up 6, you win $5. It costs $1 to play. Should you play? EV= average cost of winning + average cost of losing P(6) = 1/6 P(not 6) = 5/6 win: $5 - $1 lose: 0 - $1 EV = 1/6(5-1) + 5/6(0-1) = 4/6 - 5/6 = -1/6 On average, you lose 1/6 of a dollar.  2.2. Descriptive approaches Risk aversion/seeking Which do you choose: – a sure gain of $100 [72%] – a 50% chance to gain $200 [28%] – EV: $100 Which do you choose: – a sure loss of $100 [36%] – a 50% chance to lose $200 [64%] – EV: $100 People are risk averse (go for the sure thing) in the domain of gains and risk seeking (try not to lose) in the domain of losses (Tversky & Kahneman, 1986)  2.3. Prescriptive approaches Heuristics (& biases) “Biases in judgments reveal some heuristics of thinking under uncertainty” (Tversky & Kahneman, 1974) One reason why people make judgment errors 3 Heuristics: – Representativeness Heuristic – Availability Heuristic – Anchoring and Adjustment Heuristic  2.3. Prescriptive approaches Heuristics (& biases) Which bet would you take – which is more likely? – 5 rolls of a die = 1 1 1 1 1 – 6 rolls of a die = 1 5 3 2 5 4 (Tversky & Kahneman, 1972)  2.3. Prescriptive approaches Heuristics (& biases) Which bet would you take – which is more likely? – 5 rolls of a die = 1 1 1 1 1 – 6 rolls of a die = 1 5 3 2 5 4 – The first is more likely because it involves only 5 rolls, even though it looks unlikely (Tversky & Kahneman, 1972)  2.3. Prescriptive approaches Heuristics (& biases) Which bet would you take – which is more likely? – 5 rolls of a die = 1 1 1 1 1 – 6 rolls of a die = 1 5 3 2 5 4 – The first is more likely because it involves only 5 rolls, even though it looks unlikely Flip a coin 6 times. Which is more likely? – H H H T T T – H H T H T T (Tversky & Kahneman, 1972)  2.3. Prescriptive approaches Heuristics (& biases) Which bet would you take – which is more likely? – 5 rolls of a die = 1 1 1 1 1 – 6 rolls of a die = 1 5 3 2 5 4 – The first is more likely because it involves only 5 rolls, even though it looks unlikely Flip a coin 6 times. Which is more likely? – H H H T T T – H H T H T T – BOTH are equally likely – each flip is H or T. But one looks more random than the other. Representativeness heuristic: the subjective probability of something is determined by the extent to which it seems similar to its parent population. (Tversky & Kahneman, 1972)  2.3. Prescriptive approaches Biases & the Representativeness Heuristic Biases: – The Gambler's Fallacy – The Conjunction Fallacy – Base Rate Neglect – Ignoring the sample size  2.3. Prescriptive approaches Biases & the Representativeness Heuristic: Gambler's Fallacy Example: – In a game of chance with fair dice, a 6 has come up in the past 8 rolls. 66666666 – Should you bet that a 6 will come up again? Should you bet that a different number will be rolled, e.g., betting on a 1? (Tversky & Kahneman, 1983)  2.3. Prescriptive approaches Biases & the Representativeness Heuristic: Gambler's Fallacy Example: – In a game of chance with fair dice, a 6 has come up in the past 8 rolls. 66666666 – Should you bet that a 6 will come up again? Should you bet that a different number will be rolled, e.g., betting on a 1? – Neither! Each throw is independent of previous ones. The dice neither know nor care what happened before. – But casino owners count on gamblers trusting to the “law of averages”. Definition: – Misjudging a sequence as more “random” than another – A belief that an outcome that hasn’t happened yet is “due” (e.g., red on a roulette wheel is “due” e.g., after a run of black) (Tversky & Kahneman, 1983)  2.3. Prescriptive approaches Biases & the Representativeness Heuristic: Conjunction Fallacy 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 was an active participant in anti- nuclear demonstrations. – Rank the following statements by their probability, using 1 for the most probable and 8 for the least probable. (a) Linda works in a bookstore and takes Yoga classes. (b) Linda is active in the feminist movement. (c) Linda is a psychiatric social worker. (d) Linda is a member of the League of Women Voters. (e) Linda is a bank teller. (f) Linda is an insurance sales person. (g) Linda is a bank teller and is an active feminist. – Results: 89% of subjects ranked g as more likely than e. But g is subsumed under e – it cannot be more likely (Tversky & Kahneman, 1983)  2.3. Prescriptive approaches Biases & the Representativeness Heuristic: Base Rate Neglect Example: – Written by a high school psychologist about Tom: "Tom W. is of high intelligence although lacking in true creativity. He has a need for order and clarity, and for neat, tidy systems in which every detail fits in the appropriate place. His writing is rather dull and mechanical, occasionally enlivened by corny puns and flashes of the imagination of the sci-fi type. He has a strong drive for competence. He seems to have little feeling or sympathy for other people and does not enjoy interacting with others.” – How likely is it that Tom is currently a graduate student in: Humanities ? Computer Science ? [95%] – Results: 95% of Ss said he's more likely to be in Computer Science. (Tversky & Kahneman, 1972)  2.3. Prescriptive approaches Biases & the Representativeness Heuristic: Base Rate Neglect Example (continued): – But, when this study was run, there were 3 times as many grad students in the humanities as in Computer Science – People were ignoring base rates and relying on the representativeness heuristic. Definition: – Failing to consider the overall likelihood or frequency of something when making a decision. This is a major problem in medical diagnosis (Tversky & Kahneman, 1972; UNSW Sydney)  2.3. Prescriptive approaches Biases & the Representativeness Heuristic: Ignoring the sample size Example, assuming a 50% chance of being born male: – In Hospital A, 6 of 10 babies born (60%) were male. [Z =.63, p =.27; likely] – In Hospital B, 60 of 100 babies born (60%) were male. [Z = 2, p <.03; somewhat unlikely] – In Hospital C, 600 of 1000 babies born (60%) were male. [Z = 6.3, p 99% Nudges – Examples: Doctor appointment reminders, placement of healthy food Game Theory – Prisoners dilemma Intertemporal Choice – Marshmallow test Emotions – Lerner et al. (2004)  Conclusion How should people make decisions? What is the “best” / “optimal” decision? If you assume that “optimal”=normative, then people are biased. Those biases can be explained through models (e.g., prospect theory) or heuristics Heuristics are argued to be fast and “easy” solutions to a problem (keeping in mind our working memory limitations) People are biased to use heuristics because they are easy and often work well, but not always. Be aware.