4/26/2017. Things to Consider. Making Decisions and Reasoning. How Did I Choose to Get Out of Bed...and Other Hard Choices

Transcription

1 How Did I Choose to Get Out of Bed...and Other Hard Choices Judgments, Decision Making, Reasoning Things to Consider What kinds of reasoning traps do people get into when making judgments? What is the evidence that people sometimes make decisions that are not in their best interests? How do emotions influence decision making? Are there two ways of thinking, one fast and the other slow? Making Decisions and Reasoning Decisions The process of making choices between alternatives Reasoning The process of drawing conclusions Inductive Reasoning Reasoning that is based on observation Reaching conclusions from evidence Used to make scientific discoveries and everyday predictions 1

2 Humans are not good at estimating probabilities Often do not know objective P(Event A) Substitute subjective estimates for known P(Event A) Rely on heuristics Examples 1) From which letter string could you create more words? XUZONLCJM TAPCERHOBO 2) Which is the more likely cause of death? Breast cancer Diabetes Availability Heuristic Estimating P(A) based on how easily examples come to mind Events more easily remembered are judged as more probable Tversky and Kahneman (1973) XUZONLCJM vs. TAPCERHOBO Correlation (r =.96) between estimated and actual word production Slovic et al. (1976) Which was the more likely cause of death? Less Likely More Likely Truth Ratio % Correct Breast Cancer Diabetes 1.25 : 1 23% Lung Cancer Stomach Cancer 1.25 : 1 25% Pregnancy Appendicitis 2.00 : 1 17% Tornado Asthma : 1 42% Representativeness Heuristic Estimating P(A) based on well something matches expectations Probability that A is a member of class B based on how well the properties of A resembles properties associated with B Tversky and Kahneman (1972) In a family of three boys (B) and three girls (G), which sequence of births is more likely? B B B G G G B G G B G B 2

3 Which is more likely? P(A) P(A & B) Objective (actual) probability of the conjunction of two events (A & B) is ALWAYS less than P(A) or P(B)!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Kahneman and Tversky (1983) Conjunction fallacy (conjunction rule) Tend to estimate the conjunction of two events as more likely than one event Kahneman and Tversy (1983) Examples 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 anti-nuclear demonstrations. Linda is a bank teller. Linda is a bank teller and is active in the feminist movement. Randy is 27 years old and just graduated from art school. He loves going to museums and exhibitions and dresses flamboyantly on social occasions. Randy is a bus driver. Randy is a bus driver and paints for a hobby. Kahneman and Tversy (1983) Results (from Coglab) 7 Rating (0/Impossible - 7/Certain) Single Event Conjunction 0 Low Typical High Typical 3

4 Expected value (EV) Weighted average of all possible values a variable can take = = Expected value from $1 bet on roulette (payout is $35 for win) $1 = $ $ = $ Expected utility theory Assumes people are rational If they have all relevant information, they will make a decision that results in the maximum expected utility Desirable outcomes Outcomes that are in the person s best interest Advantages Specific procedures to determine the best choice Rational Disadvantages Not necessarily money, people find value in other things Many decisions do not maximize the probability of the best outcome Denes-Raj & Epstein (1994) 4

5 Expected utility (EU) Weighted average of utilities of each outcome associated with an alternative = Must assign probability and utility to each possible outcome Expected utility model example Must choose between PSYC 224 and PSYC 235 Assume each course has two possible outcomes PSYC 224 p =?? Easy course, get good grade (u =??) p =?? Boring, becomes uninteresting (u =??) Start p =?? Challenging, helps grad school (u =??) PSYC 235 p =?? Too difficult, becomes unfun (u =??) Expected utility model example p =.75 Easy course, get good grade (+10) PSYC 224 p =.25 Boring, becomes uninteresting (-10) Start p =.70 Challenging, helps grad school (+25) PSYC 235 p =.30 Too difficult, becomes unfun (-15) 5

6 Expected utility model example 224 = = =5 235 = = =13 EU(PSUC 235) > EU(PSUC 224) Should choose PSYC 235 Prospect Theory Make decisions based on framing of potential losses and gains Losses and gains are evaluated by heuristics, not just probabilities Risk aversion Used when problem is stated in terms of gains Avoid risky, uncertain gains Example: Potential Outcomes Choice A 80% chance +$4000 and 20% chance ±$0 Choice B Expected Value (EV) +$3200 +$3000 % Choosing ~20% ~80% 100% chance of +$3000 6

7 Risk seeking (loss aversion) Prefer risky actions when there is a potential for loss Used when problem is stated in terms of losses Example Choice C Potential Outcomes 80% chance of -$ % chance of ±$0 Choice D Expected Value (EV) -$3200 -$3000 % Choosing 92% 8% 100% chance -$3000 Proportion Choosing Risky Decision Risky Decisions CogLab Expected Results Gain Loss Question Frame Proportion Choosing Risky Decision Your Results Gain Loss Question Frame Framing effects Decisions are influenced by how a decision is stated The nation is preparing for the outbreak of a disease that is expected to kill 600 people. Two alternative programs to combat the disease have been proposed by public health officials; which would you support/choose? (A) Two hundred people will be saved. (B) There is a one-third probability that 600 people will be saved, and a 2/3 probability that no one will be saved. The nation is preparing for the outbreak of a disease that is expected to kill 600 people. Two alternative programs to combat the disease have been proposed by public health officials; which would you support/choose? (C) Four hundred people will die. (D) There is a one-third probability that nobody will die, and a 2/3 probability that 600 people will die. 7

8 Tversky and Kahnemann (1981) When situations are framed in terms of gains, people tend toward a risk-aversion strategy When situations are framed in terms of losses, people tend toward a risk-taking strategy Monty Hall Let s Make a Deal! 8

9 The question / paradox: Does switching matter? Should you stay with your original choice or switch to the other, unopened door? P(Win Stay) P(Win Switch) Marilyn vos Savant P(Win Stay) P(Win Switch) Your Decision after DOOR 1 DOOR 2 DOOR 3 MH Opens Door Your 1/3 1/3 1/3 Switch Stay p(situation) P(Win) Selection Door 1 You M.H. Lose Win 1/6 1/3 Door 1 You M.H. Lose Win 1/6 Door 2 You M.H. Win Lose 1/6 Door 2 You M.H. Win Lose 1/6 2/3 Door 3 M.H. You Win Lose 1/6 Door 3 M.H. You Win Lose 1/6 P(Win Stay) P(Win Switch) P(Win) = 1/3 P(Win) = 2/3 9

10 CogLab Results Frequency (nno.) Percentage Wins (%) Wins Losses Percentage wins Stay Switch 0 Decision After Choosing First Door 10