Computational Models of Thought and Behavior

Fall 2020

Course site for MGT 451A, with a list of class meetings and assignments.

separator

In this course, we review computational models of thought and behavior from across the behavioral and social sciences, including economics, psychology, evolutionary biology, network science, information systems, and sociology. The course begins by studying individual thought and behavior, from predicting the future to taking risks. The course proceeds to study group thought and behavior, from social influence to the design of voting mechanisms. Each class covers a new topic in the study of thought and behavior and introduces the range of empirical and theoretical computational approaches to studying that topic.

separator

Class meetings

A whirlwind tour of thought and behavior, together and apart.

Monday, 8/26

Topics:

  • Overview of class
  • Small-group discussion to spark interest
  • Review this syllabus

Readings:

  • This syllabus
  • “The Standard Equipment” by Steve Pinker
  • NCM Chapter 1

Part I: Computational models of individual thought and behavior.

On predicting the future.

Day, Date

Topics:

  • The Anthropic Principle, in
  • Basics of Bayesian inference
  • “Optimal predictions in everyday life” by Griffiths & Tenenbaum.
  • The end-of-history illusion
  • The recency illusion
  • Retrofuturism

Readings:

  • The Anthropic Principle by J. Richad Gott
  • Retrofuturism subreddit (https://www.reddit.com/r/RetroFuturism/top/?t=all)

Assignment #1: In this assignment, you will replicate the experiment detailed in the “Optimal predictions…” reading.

On reasoning about what might have been.

Day, Date

Topics:

  • Causal graphical models
  • Measures of simplicity in causal explanation
  • The survivorship bias

On seeking and avoiding risk.

Day, Date

Topics:

  • Kahneman and Tversky and the history of behavioral economics
  • Risk aversion
  • The certainty effect
  • Expected Value Theory
  • Expected Utility Theory

Assignment #2: In this assignment, you will implementing Expected Value Theory and Expected Utility Theory using the Python language and demonstrate a choice anomaly.

On gaining and losing.

Day, Date

Topics:

  • Prospect theory
  • “When less is more” and counterfactual reasoning
  • Anchoring and adjustment

On choosing between now and later.

Day, Date

Topics:

  • Intertemporal choice
  • Hyperbolic vs. exponential discounting
  • Cooperating with the future

Assignment #3: In this assignment, you will measuring your own temporal discounting function and determine whether it is better described as a hyperbolic or exponential function.

On being rational.

Day, Date

Topics:

  • Marr’s three levels and definitions of rationality
  • The Allais paradox
  • Ellsberg paradox
  • The Law of Small Numbers
  • Gambler’s Fallacy

On exploring and exploiting.

Day, Date

Topics:

  • Multiarmed bandit problems
  • The Gittins index
  • The Upper Confidence Bound algorithm
  • Epsilon-greedy strategies for solving multiarmed bandit problems
  • Thompson sampling
  • From multiarmed bandits to contextual bandits

On deciding under incomplete information.

Day, Date

Topics:

  • Markov Processes
  • Markov Decision Processes
  • Partially Observable Markov Decision Processes

Assignment #4: In this assignment, you will implement a Partially Observable Markov Decision process using the Python programming language.

On thinking, fast and slow.

Day, Date

Topics:

  • Drift diffusion models of choice
  • System I and System II thinking
  • The idea of time as a resource

On summarizing one’s experience.

Day, Date

Topics:

  • Peak-end rule
  • Extension neglect
  • Duration neglect
  • Representativeness heuristic

On having limits.

Day, Date

Topics:

  • Bounded rationality
  • The concept of attentional limits
  • Availability heuristics
  • Base-rate neglect
  • Conjunction fallacy
  • Resource rationality

Assignment #5: In this assignment, you will test some of your own perceptual and attentional limits and consider various ways to quantify them.

On knowing thyself.

Day, Date

Topics:

  • Metacognition
  • Dunning-Kruger effect
  • Curse of knowledge

On having preferences.

Day, Date

Topics:

  • Preferences
  • Implicit Associations
  • Multi-objective scalarization
  • Multi-objective optimization

Assignment #6: In this assignment, you will take an online version of the Implicit Association Test.

On wishing it were so.

Day, Date

Topics:

  • Motivated reasoning
  • Survivorship bias
  • Selection biases

“But I like it / Because it is bitter / And because it is my heart.”

Day, Date

Topics:

  • The Endowment Effect
  • The Ikea Effect
  • Not Invented Here disorder
  • The Illusion of Control

On the troubled history of intelligence.

Day, Date

Topics:

  • What is intelligence?
  • The troubled history of studying intelligence
  • The foundations of eugenics
  • Measuring and modeling “general” intelligence (G)
  • The concept of multiple intelligences
  • The Intelligence Quotient (IQ)

Midterm exam.

Part II: Computational models of collective thought and behavior.

On evolving over time.

Day, Date

Topics:

  • Is evolution actually survival of the fittest?
  • The Iterated Prisoner’s Dilemma (IPD)
  • IPD on structured networks

Readings:

  • NCM Chapter 6

Assignment #7: In this assignment, you will write an evolutionary IPD tournament in Python.

On the wisdom of the crowd.

Day, Date

Topics:

  • Francis Galton and the voice of the people
  • The Delphi Method
  • The Bayesian Truth Serum

Readings:

  • Vox Populi
  • Bayesian Truth Serum paper

Assignment #8: In this assignment, you will replicating the wisdom of the crowds effect from the Galton reading.

On designing good mechanisms.

Day, Date

Topics:

  • Algorithmic Game Theory
  • Game theory vs. mechanism design
  • Algorithmic mechanism design
  • The Price of Anarchy
  • Mechanisms for Voting
  • Mechanisms for Auctions

On collective intelligence.

Day, Date

Topics:

  • Schooling in fish
  • Social learning
  • Collective intelligence
  • Crowdsourced markets

On recursive social reasoning.

Day, Date

Topics:

  • Theory of mind
  • Schelling coordination games
  • Keynesian beauty contests
  • El Farol Bar Problem
  • The minority game
  • Limits to recursive reasoning

On social influence.

Day, Date

Topics:

  • Conformity bias
  • Prestige bias
  • Herding in crowds
  • Information cascades
  • Ingroup bias
  • Fundamental attribution error

On being fair.

Day, Date

Topics:

  • Algorithmic fairness
  • Cake cutting
  • Fair division

Assignment #9: In this assignment, you will studying the fairness of an AI system.

On being altruistic.

Day, Date

Topics:

  • The Dictator Game
  • The Ultimatum Game

On remembering together.

Day, Date

Topics:

  • Collective memory
  • Transactive memory
  • Interactive cueing
  • Forgetting on networks

Assignment #10: In this assignment, you will replicate the effect of interactive cueing with a friend.

Final exam.

separator

Assignments

separator