Computational Modeling Approach

to Decision Making

 

 

1. Measurement of Preference

 

·        Decision making theories begin with the concept of a preference relation.

 

A, B, C are alternatives (or options)

       Gambles, Cars, Jobs, Houses, Medical Treatments

A ³p B means A is preferred or indifferent to B

 

·        Preference relations can be measured by

          Choice   (choose between A or B)

          Certainty Equivalents (what is the dollar equivalent of each option)

          Ratings (rate how strongly you like each option on a 10 point scale)

 

·        Different Measures of Preference do not always yield the same order

Producing preference reversals

 

e.g.,

Gamble A:  .95 chance of winning $4  vs. nothing

Gamble B:  .60 chance of winning $16 vs. .4 chance of losing $8

 

Choice Frequency A > Choice Frequency B

Certainty Equivalent for B > Certainty Equivalent for A

(see Slovic & Tversky, 1993, for a review)

 

·        Most theorists believe that choice is the most basic measure of preference (see Luce, 2000)

 


2. Conflict and the Probabilistic Nature of Preference

 

         Suppose a person is given a choice between two options that are approximately equal in weighted average value, inducing some type of conflict.

 

         The same pair of options is presented on two different occasions.

 

         The probability of making an inconsistent choice is .33. In other words, the person changes his or her mind 1 out of 3 times! (see, e.g., Starmer, 2000)

 

         The test – retest (within one week) correlation for selling prices is generally below .50 (less than 25% predictable across time). (Hershey & Schoemaker, 1989)

 

         This is a ubiquitous property of human behavior, but standard utility theories consider it an irrational aspect of human choice.

 

 

3. Biological and Evolutionary Explanations for the Probabilistic Choice

 

·        Exploratory Behavior

o       We need to continuously learn about uncertain probabilities of payoffs in a changing, non-stationary environment.

 

·        Unpredictable Behavior

o       We do not want our competitors to be able to perfectly predict our behavior and use this to take advantage of us.

 

·        Dynamic Motivational Systems

o       Our needs or goals change over time like hunger, thirst, sex

 


4. Psychological Explanations for Probabilistic Choice

 

·        Fundamental Preference Uncertainty

o       We have fuzzy beliefs and uncertain values.

 

·        Constructive Evaluations

o       We need to construct evaluations online, and the frame may change, and attention may fluctuate.

 

·        Changing Strategies

o       Using different choice rules can change preferences

 


5. Implications for Standard Utility Theory

 

·        Suppose we assume: Choose A over B à A ³p B à u(A) > u(B)

o       What problems does this generate?

 

·        MaCrimmon (1968) asked 38 business managers to respond to 3 sets of choices, and 8 managers exhibited intransitivity’s. Should we reject utility theory?

 

·        Absolutely not (e.g.,says Luce, 2000) these are just errors. After all, the nature of choice is probabilistic.

 

·        Thus, standard utility theorists define

 A ³p B à Pr[ A | {A,B} ] ³ .50

 

·        In the end, standard utility models are actually founded on probabilistic choice assumptions. Axioms must be tested using statistical models.

 

·        But why is a .00002 change in probability from .49999 to .50001 more important than a .48 change in probability from .51 to .99 or .49 to .01?

 

·        A model that accounts for the entire continuous range of probabilities is superior to one that only accounts for two categories [0,.5) vs (.5, 1] of probabilities.

 


6. Decisions take time

 

·       Decision time is systematically related to choice probability

o     Petrusic and Jamieson (1978)

o     Dror, Busemeyer, & Baselo al (1999)

 

·       Choice probabilities become more extreme with longer deliberations

o     Simonson (1989) compromise effect

o     Dhar (2000) attraction effect

 

·       Preferences can be reversed under time pressure

o     Edlund and Svenson (1993)

o     Diederich (2000)

 

·        Preferences are dynamically inconsistent (Plans are not followed)

o       Ainslie (1975)

o       Busemeyer et al. (2000)

o       Trope et al (2002)

 

7. Goals of Computational Models of Choice

 

·        Explain how conflicts are resolved

o       the deliberation process described by William James

 

·        Account for the entire continuous range of choice probabilities [0,1]

o       Not simply categorize whether they are above or below 50%

o       Explain paradoxical choice behavior

 

·        Account for other manifestations of choice

o       Choice response time

o       Confidence Ratings

 

·        Account for other manifestations of preference

o       Certainty equivalents

o       Buying or selling prices

 

·        Explain the origins of weights and values

 

·        Build on principles from both cognitive psychology and neuro-psychology

 

·        Examples

o       Decision Field Theory (Busemeyer & Townsend, 1993)

o       Neural Computational Model of Usher & McClelland (2002)

o       Constraint Satisfaction model of Guo and Holyoak (2002)