Testing Models of Stochastic Dominance Violations Michael H. Birnbaum Decision Research Center California State University, Fullerton.

Slides:



Advertisements
Similar presentations
Cognitive Modelling – An exemplar-based context model Benjamin Moloney Student No:
Advertisements

New Paradoxes of Risky Decision Making that Refute Prospect Theories Michael H. Birnbaum Fullerton, California, USA.
PSME M1 Economic Growth Tutorial.  Introduction ◦ Review of Classic Solow Model ◦ Shortfalls of Solow ◦ Human Capital Accumulation ◦ Convergence Theory.
Stat 112: Lecture 7 Notes Homework 2: Due next Thursday The Multiple Linear Regression model (Chapter 4.1) Inferences from multiple regression analysis.
Effort Discounting of Exam Grades Heidi L. Dempsey, David W. Dempsey, & Arian Ward Jacksonville State University.
Among those who cycle most have no regrets Michael H. Birnbaum Decision Research Center, Fullerton.
This Pump Sucks: Testing Transitivity with Individual Data Michael H. Birnbaum and Jeffrey P. Bahra California State University, Fullerton.
All About Dice Dice have been around for at least 5000 years and are used in many games. Knucklebones of animals, which are approximately tetrahedral.
1 Upper Cumulative Independence Michael H. Birnbaum California State University, Fullerton.
1 Lower Distribution Independence Michael H. Birnbaum California State University, Fullerton.
True and Error Models of Response Variation in Judgment and Decision Tasks Michael H. Birnbaum.
Evaluating Non-EU Models Michael H. Birnbaum Fullerton, California, USA.
Who are these People Who Violate Stochastic Dominance, Anyway? What, if anything, are they thinking? Michael H. Birnbaum California State University, Fullerton.
McGraw-Hill Ryerson Copyright © 2011 McGraw-Hill Ryerson Limited. Adapted by Peter Au, George Brown College.
© red ©
Testing Lexicographic Semi- Order Models: Generalizing the Priority Heuristic Michael H. Birnbaum California State University, Fullerton.
Testing Heuristic Models of Risky Decision Making Michael H. Birnbaum California State University, Fullerton.
1 A Brief History of Descriptive Theories of Decision Making Kiel, June 9, 2005 Michael H. Birnbaum California State University, Fullerton.
Some New Approaches to Old Problems: Behavioral Models of Preference Michael H. Birnbaum California State University, Fullerton.
1 Distribution Independence Michael H. Birnbaum California State University, Fullerton.
1 Upper Tail Independence Michael H. Birnbaum California State University, Fullerton.
1 Upper Distribution Independence Michael H. Birnbaum California State University, Fullerton.
Ten “New Paradoxes” Refute Cumulative Prospect Theory of Risky Decision Making Michael H. Birnbaum Decision Research Center California State University,
Violations of Stochastic Dominance Michael H. Birnbaum California State University, Fullerton.
Testing Critical Properties of Models of Risky Decision Making Michael H. Birnbaum Fullerton, California, USA Sept. 13, 2007 Luxembourg.
Ten “New Paradoxes” Refute Cumulative Prospect Theory of Risky Decision Making Michael H. Birnbaum Decision Research Center California State University,
New Paradoxes of Risky Decision Making that Refute Prospect Theories Michael H. Birnbaum Fullerton, California, USA.
1 The Case Against Prospect Theories of Risky Decision Making Michael H. Birnbaum California State University, Fullerton.
Testing Transitivity (and other Properties) Using a True and Error Model Michael H. Birnbaum.
Web-Based Program of Research on Risky Decision Making Michael H. Birnbaum California State University, Fullerton.
Web-Based Program of Research on Risky Decision Making Michael H. Birnbaum California State University, Fullerton.
1 A Brief History of Descriptive Theories of Decision Making: Lecture 2: SWU and PT Kiel, June 10, 2005 Michael H. Birnbaum California State University,
REGRESSION MODEL ASSUMPTIONS. The Regression Model We have hypothesized that: y =  0 +  1 x +  | | + | | So far we focused on the regression part –
1 Gain-Loss Separability and Reflection In memory of Ward Edwards Michael H. Birnbaum California State University, Fullerton.
I’m not overweight It just needs redistribution Michael H. Birnbaum California State University, Fullerton.
1 Ten “New Paradoxes” of Risky Decision Making Michael H. Birnbaum Decision Research Center California State University, Fullerton.
1 Gain-Loss Separability Michael H. Birnbaum California State University, Fullerton.
T-Tests Lecture: Nov. 6, 2002.
Is there Some Format in Which CPT Violations are Attenuated? Michael H. Birnbaum Decision Research Center California State University, Fullerton.
1 Lower Cumulative Independence Michael H. Birnbaum California State University, Fullerton.
Stochastic Dominance Michael H. Birnbaum Decision Research Center California State University, Fullerton.
Web-Based Program of Research on Risky Decision Making Michael H. Birnbaum California State University, Fullerton.
Testing Transitivity with Individual Data Michael H. Birnbaum and Jeffrey P. Bahra California State University, Fullerton.
1 Restricted Branch Independence Michael H. Birnbaum California State University, Fullerton.
Learning Chapter 18 and Parts of Chapter 20
Presidential Address: A Program of Web-Based Research on Decision Making Michael H. Birnbaum SCiP, St. Louis, MO November 18, 2010.
Gaussian process modelling
Behavior in the loss domain : an experiment using the probability trade-off consistency condition Olivier L’Haridon GRID, ESTP-ENSAM.
Decision making Making decisions Optimal decisions Violations of rationality.
The Theory of Consumer Choice
Sequential Expected Utility Theory: Sequential Sampling in Economic Decision Making under Risk Andrea Isoni Andrea Isoni (Warwick) Graham Loomes Graham.
Ellsberg’s paradoxes: Problems for rank- dependent utility explanations Cherng-Horng Lan & Nigel Harvey Department of Psychology University College London.
Buying and Selling Prices under Risk, Ambiguity and Conflict Michael Smithson The Australian National University Paul D. Campbell Australian Bureau of.
Testing Transitivity with a True and Error Model Michael H. Birnbaum California State University, Fullerton.
We report an empirical study of buying and selling prices for three kinds of gambles: Risky (with known probabilities), Ambiguous (with lower and upper.
Chapter 7: Sampling Distributions Section 7.1 How Likely Are the Possible Values of a Statistic? The Sampling Distribution.
Can a Dominatrix Make My Pump Work? Michael H. Birnbaum CSUF Decision Research Center.
Chapter 9: Introduction to the t statistic. The t Statistic The t statistic allows researchers to use sample data to test hypotheses about an unknown.
Statistics for Business and Economics Module 1:Probability Theory and Statistical Inference Spring 2010 Lecture 4: Estimating parameters with confidence.
Chapter 9 Estimation using a single sample. What is statistics? -is the science which deals with 1.Collection of data 2.Presentation of data 3.Analysis.
Chapter 9 Estimation and Confidence Intervals. Our Objectives Define a point estimate. Define level of confidence. Construct a confidence interval for.
Bagging and Random Forests
Grading on a curve, and other effect of group size on all-pay auctions
Estimating with Confidence
Colours.
Dopamine Reward Prediction Error Responses Reflect Marginal Utility
What Color is it?.
Inactivation of Medial Frontal Cortex Changes Risk Preference
Perceptual Classification in a Rapidly Changing Environment
New Paradoxes of Risky Decision Making that Refute Prospect Theories
Presentation transcript:

Testing Models of Stochastic Dominance Violations Michael H. Birnbaum Decision Research Center California State University, Fullerton

2 How to Explain violations of Stochastic Dominance? Violations of SD refute a large class of theories including CPT. Which model, TAX or RAM, makes more accurate predictions? Series of 5 studies designed to investigate systematic manipulations.

3 Stochastic Dominance However, it is possible to find 70% violations of this property among undergraduates.

4 Violation of Stochastic Dominance Gamble AGamble B 90 reds to win $96 05 blues to win $14 05 whites to win $12 85 reds to win $96 05 blues to win $90 10 whites to win $12 70% of undergrads choose B

5 Same Choice in Split Form Gamble A SplitGamble B Split 85 reds to win $96 05 greens to win $96 05 blues to win $14 05 whites to win $12 85 reds to win $96 05 greens to win $90 05 blues to win $12 05 whites to win $12 90% choose C over D

6 Studies of SD: Models and Heuristics Do people violate SD by simply averaging the consequences and ignoring probabilities? Is RAM or TAX more accurate in predicting when violations ARE or ARE NOT observed?

7 Models Require No New Parameters RAM and TAX models are used with their parameters from previous data to make predictions to the new data. One “new” parameter, logistic spread parameter, estimated from previous data used to predict choice percentages.

8 Logistic Choice Model Parameter, , was chosen to fit the previous result that G- is chosen over G+ on 70% of choices.

9 G– = ($96,.85 – r; $90,.05; $12,.1 + r)

10 Data Violate the Consequence Averaging Heuristic Data from two studies with n = 330 show that participants respond to changes in probability. Rules out the consequence averaging heuristic. RAM violates probability monotonicity when probability is transferred from highest to middle branch. Tested in next study (n = 428).

11 Probability Monotonicity If probability is shifted from a branch leading to a lower-valued consequence to a branch leading to a better consequence, the utility of the gamble should be increased. RAM/TAX satisfy PM in two-branch gambles, but not in three-branch gambles.

12 Study 3 (n = 428) Choices constructed from the following: G+ = (96,.9; 14,.05; 12,.05) versus G- = (96,.85 - r; 12,.05 + r; 12,.05) Probability Monotonicity implies that the probability of choosing G- over G+ should decrease as r is increased.

13 Predictions The RAM model implies violations of PM as probability is shifted from the highest branch to the middle branch. TAX implies a very small effect, with a slight increase.

14 G– = ($96,.85 – r; $90,.05 + r; $12,.1)

15 Results The data show an increase, contrary to RAM as fit to previous data. The increase is larger than predicted by TAX, which predicts virtually no effect of the manipulation.

16 Consequence Counting Heuristic Perhaps people count how many branches of a gamble give higher consequences and choose the gamble with the greatest number of higher consequences. We can alter the consequences so that G+ has two branches with higher consequences.

17 SD Study 4: Consequences 90 black win $97 05 yellow win $15 05 purple win $13 85 red to win $95 05 blue to win $91 10 white to win $11 Predictions of TAX: 70% for ($97, $13) 68% for ($95, $11). Observed: 72% and 68% n = 315

18 SD Study 4: middle Branch 90 red win $96 05 blue win $14 05 white win $12 85 red win $96 05 blue win $70 10 white win $12 Predictions of RAM and TAX Are 64% and 60%, respectively. Observed is 70%, n = 315.

19 Effect of Middle Branch

20 Results: Study 4 The effect of the middle branch was more closely predicted by TAX than RAM. Reducing the other two branches by small amounts, to vary the number of branches favoring the dominant gamble had very small effects, approximated by TAX. This manipulation had less effect than predicted.

21 SD Study 5: All 3 conseqs. 90 black win $97 05 yellow win $15 05 purple win $13 85 red win $90 05 blue win $80 10 white win $10 Predictions of TAX and RAM are 63% and 50%, respectively. Observed is 57%*, n = 394

22 Summary: 5 Studies of SD, 1467 participants People respond to changes in probability, contrary to counting heuristic. Both RAM and TAX can violate probability monotonicity, data closer to TAX than RAM. (* more) People respond to changes in consequences, but not extremely.

23 To Save Heuristics These data are not consistent with the consequence averaging heuristic, nor are they consistent with the contrast counting heuristic. To save heuristics, we might argue that they work only in some people some of the time. We can also change the heuristic to a different one in response to data.

24 Next Program: Demographic Correlates Does the rate of violation of SD relate to education, particularly training in decision- making? Next program will review web-based research that sought to recruit highly educated participants. Indeed, PhDs have lower rates of violation, but violations are still substantial, and would refute CPT.

25 For More Information: Download recent papers from this site. Follow links to “brief vita” and then to “in press” for recent papers.