Simulation and Modeling in the Social and Policy Sciences

http://ppol225.danryan.us
Mills College, Fall 2013
Lecture Tu-Th 2:30-3:45PM AHR 2001 Lab : Th 4-6:00PM STR 14
Instructor: Dan Ryan | danryan at mills dot edu | @djjr | 510-430-3242 | Vera Long 105
Office Hours M, T, W by Appointment http://officehours.danryan.us

*Quick Jumps:** Course Intro | Flowcharts | Decision Models | Difference Equations | Stock and Flow Models (system dynamics) | Markov Models | Sorting and Peer Effects | Aggregation | Cellular Automata | Tipping Models | Diffusion | Coordination and Cooperation | Biases | Path Dependence | Mechanism Design | The Wisdom of Crowds |

+Graphic

Preliminaries

Description

Extended Description

+ Show

Plan of the Course. This course is the hybrid daughter of the course I have taught at Mills for the past several years and one offered on Coursera by Scott E. Page (University of Michigan). His course is very similar to what this course started as some years ago and he's an expert in the field and experimentation is good so we are trying something new.

The course will begin with several different authors' comments on the question of what models are and what they are good for. These perspectives are complementary rather than duplicative. Triangulation is the intended effect. We then move into flow charts as models of deterministic processes. We'll see flow charts play three roles: descriptive (an aid to understanding processes), prescriptive (an aid to implementing a protocol), and developmental (as an aid to designing processes). In conjunction with this we will introduce the concepts of black boxing, top-down design, and stepwise refinement.
Next we will turn our flow charts on their side and model non-deterministic decision processes, introducing decision models and decision trees. By the time we finish with this topic we will have introduced methods for determining the value of information used in making decisions and some techniques for dealing with imperfect information when making decisions.

Before proceeding we want to pause for a moment and introduce a very simple and very general tool called "difference equations." We will come to know a whole family of phenomena that can be modeled using difference equations and we will meet the idea of different ways that a system can evolve, either moving toward an equilibrium or not. Along the way we will encounter another manifestation of this approach called "stock and flow" models and introduce the idea of positive and negative feedback.

After stock and flow we'll introduce a related model in which entities "move" in a system: Markov models.

The models we will have seen at this point are either individual (decisions) or feature things that are already aggregated. One of the really useful aspects of models and simulations, though, is when we bring these together and ask how individual level actions aggregate to yield collective outcomes. As you might imagine, what makes this interesting is that the whole is often the non-obvious sum of the parts. The next several sections of the course explore this in various ways.

We start with thinking about how to model phenomena in which people's behavior depends on the behavior of those around them, introducing the general idea of "micro-motives" and "macro-behaviors" and highlight the important insight that individual choices can give rise to contrary collective effects. (Sorting and Peer Effects, Aggregation, Tipping).

The next section of the course begins with some thinking about a fundamental question: what kind of thing is an actor? That is, a human actor. It's their behavior, after all, that we've been talking about aggregating. Scott Page's section, "Thinking Electrons," will introduce us to rational actors and rule following actors — both of which we have been using all along — and then complicate matters with some insights from behavioral economics. From here we will move to a consideration of how to model coordination and cooperation (Coordination and Culture, Prisoners' Dilemma and Collective Action).

The models we considered in the last sections mostly omitted the question of history. In the simple models, at least, everyone acts at the same time and no one remembers. But is history something we can always ignore? Here we will look at "path dependence" as an element of modeling.

But what can we do with these things? As policy analysts, activists, or social scientists, we would like to be able to predict what will emerge from a given situation and or design institutions that will foster the emergence of a desired outcome. The last section of the course considers "applying" model thinking to our work.

Books

+ Show

1. Stokey, Edith and Richard Zeckhauser. A Primer for Policy Analysis.

Optional

1. Schelling, Thomas. Micromotives and Macrobehavior.
2. Lave, Charles A. and James G. March An Introduction to Models in the Social Sciences

Requirements

+ Show

Experience has shown that the best predictor of success in this course is keeping up with the work and repetition. To better incentivize this, a significant proportion of the semester grade will be for doing work on-time. Relatedly we will have more smaller exams rather than fewer larger exams. On the lab side, we'll "award" points for participation (showing up and handing things in) but we'll also have an exam specifically on the skills we learn in lab. The exams may be either take-home or in-class at the instructor's discretion.

With allowances for adjustments to take into account changes in course schedule and plans, the final grade will be calculated on the following basis:

	"Points"
On-time submission of problem sets/quizzes	20
Professional appearance of problem sets/quizzes	20
Exams 1,2,3,and 4 (total)	20+20+20+20
Final Exam	40
Total Possible	160

A grades are likely in the 80-90% range, B 70-80% and C 60-70% range.

Lab attendance is should be understood to be mandatory and class attendance is taken for granted. Unexcused absences after two classes or one lab may result in a reduction of up to 2/3 of a letter grade (though, usually the penalty is self-executing as other work tends to suffer).

Academic Integrity

+ Show

Customary academic standards academic integrity (including proper bibliographic citation) apply.  It is your responsibility to know what these are and to follow them.  Collaborative learning is encouraged, but work that is submitted under your name as a demonstration of your skills and competence must represent YOUR work.  Plagiarism, as defined under the Mills College Honor Code, will be cause for, at a minimum, a failing grade in this course. Please consult with instructor if you have any questions, or even the slightest doubt, about how to follow these requirements.  When in doubt, cite.

Accommodations

+ Show

Every effort will be made to make this class accessible for students regardless of disability. Students with needs for accommodation should contact for Students with Disabilities (Cowell Building, x2130) and inform the instructor in order for access to be arranged adequately and promptly.

Schedule

NOTE.  The schedule below indicates when I expect we will be reading/discussing the readings in the course text book.  Other recommended readings (and even the occasional required reading) may be added to the online syllabus from time to time (with plenty of notice and instructions on how to obtain the readings).  We may fall behind this schedule from time to time.  I will try to always mention what we are reading next at the end of each session.  In any case, though, it is the student's responsibility to (1) read/view things on time, and (2) keep abreast of where we are in the course schedule.

Introduction

+ Show Section Description

Lectures and Reading

Week 0 Thursday August 29

Class: Course Introduction — What are models and why do we model?

1. Instructor Notes
2. Introduction to PPOL225
3. Introduction to Coursera & Scott E. Page (SEP)
4. Lecture 1.1: Why Model? (8:52)
5. How to write up problem solutions

Lab Excel Skills

1. Instructions
2. Deliverable: TBA
3. Excel skill list

After Thursday's classes:

1. Get a Coursera account and sign up for Model Thinking.
2. View the remaining lectures in Section 1 of Model Thinking,  (about 1 hour)
   • Page, Lecture 1.2: Intelligent Citizens (11:30)
   • Page, Lecture 1.3: Thinking More Clearly (10:50)
   • Page, Lecture 1.4: Using and Understanding Data (10:15)
   • Page, Lecture 1.5: Using Models to Decide, Strategize, and Design (15:25)
3. Read ch. 2 "Types of Models" in Stokey and Zeckhauser*.
4. Read The Model Thinker: Prologue, Introduction and Chapter 1
5. Read Why Model? by Joshua Epstein
6. (optional) Read pp. 11-43 of Schelling's Micromotives and Macrobehaviors.

DUE before 11 p.m. Sunday September 1

REVIEW QUIZ
WORK REPORT

FLOW CHARTS

Flow Charts

Flow Chart Skills

+ FC01 Create flow chart from textual description

MTC
Problems
(Edit)

+ FC02 Stepwise Refinement

Use flowchart to demonstrate stepwise refinement and black-boxing (Edit)

+ FC03 Flow charts with time and division of labor

Create and interpret flow chart that includes time and divisions of labor (Edit)

+ FC04 Diagnose bad Flowcharts

Offer critique and corrections to a poorly formed flowchart (Edit)

+ FC05 Issues and Caveats

Identify and comment on concerns and caveats (Edit)

+ FC06 Software Tools

Draw flow charts in Microsoft Excel, Word or other software (Edit)

+ Section Description

TO DO

CLASS 1

BEFORE Class
READ

1. Ryan. Ryan on Flowcharts I
2. HCI Consulting. An overview (optional)

DO

1. Univ Plymouth, UK. Flow Charts for Simple Tasks: Tutorial with exercises
2. Univ Plymouth, UK. Flow Charts for Classification: Tutorial with exercises
3. problems 71, 72, 73, 74, and 75 (in problem notebook)

IN Class

1. Review of rules; Concept and utility; Deterministic models;
2. Problems 76 and 83
3. Instructor Notes

CLASS 2 and LAB

BEFORE Class

1. Read Ryan on Stepwise Refinement
2. Do problems 79 and 80.
3. Read Flow charts with time and division of labor
4. Do problem 87.

Before Lab

1. Problems 88, 89, 90

IN Class

We will work on problems 76, 83, 85, 86, 87, and 91 during this class session.

IN Lab
Instructions
Deliverables

DUE on module deadline

1. Section Quiz Flowcharts

1. Weekly Work Report

Tu 10-Sept to Th 19-Sept

Decision Models

+ Section Description

Understanding how decisions are made and how to make decisions is a useful skill for the policy analyst. Sometimes we want to know how to take what we know or think we know and put it together to reach a decision. Sometimes we want to know what we would need to know (and how much we are willing to work to find out) in order to make a decision. Sometimes we want to understand how members of a polity are making decisions.

This section of the course introduces models of how people make decisions : multi-criterion decision making, spatial models of decision making, and decision trees. The last section will introduce the idea of the value of information in decision making.

Readings and Videos

• Stokey & Zeckhauser, ch. 12 "Decision Making"
• 4.1 Introduction to Decision Making (5:37)
• 4.2 Multi-Criterion Decision Making (8:18)
• SEP Spatial Models (5pp)
• 4.3 Spatial Choice Models (11:08)
• SEP Multi Criterion Decision Making Case Study
• 4.4 Probability: The Basics (10:06)
• SEP Decision Theory
• 4.5 Decision Trees (14:38)
• 4.6 Value of Information (8:41)

Week 2 Tuesday September 10

BEFORE class

1. Watch lecture 4.1, "Introduction to Decision Making," (5:37) for an overview of decision models
2. Watch 4.2 "Multi-Criterion Decision Making" (8:18)
3. Watch 4.3 "Spatial Choice Models" (11:08)
4. (optional) Read multi-criterion decision models and the MCRI. (~28p)
5. (optional) Read lightly: spatial preferences (~5p)
6. Quiz on Page lectures

1. Watch lecture 4.4, "Probability: The Basics" (10:06)
2. Read the first 8 pages of Decision Theory with focus on section 1.2, pp 5-8 on expected value
3. Read one hour into Stokey-Zeckhauser chapter 12 or at least to page 216 (DL)
4. Refresher quiz on probability.
5. If your probability skills feel like they need some work, consult online review/tutorial sites such as CIMT (Lancaster, UK): Basic Probability of One Event, Finding Outcomes With Two Experiments, Finding Probabilities Using Lists Of Outcomes, Multiplication Law For Independent Events

Class: Introduction to Decisions and Review of Probability
We will introduce decision trees, choice and chance nodes, and review probability and expected value. Note: class will begin with a very short probability diagnostic quiz.

After Class Preliminary Decisions Quiz

Week 2 Thursday September 12

Before Class

1. Watch lecture 4.6, "Value of Information" (8:41)
2. Re-read pp. 206-21, especially 219-221 in S&Z
3. Problems 93, 95, 96, 97, 98, 101

Class: The Value of Information
Decisions under uncertainty with testing

Lab 4 Decision Trees
Excel skills: formatting, string formulas, conditional formatting, spinners, joining cells, borders

• Instructions (TO BE REVISED)
• Deliverables TBA

Week 3 Tuesday September 17

BEFORE Class

1. Re-read all of the S&Z material on decision trees
2. Re-read SEP material on decision trees as necessary.
3. Be prepared to work on problems at end of SEP Decision Theory in class.

Class Decision Trees and Decision Strategies
Problems from Decision Theory

Quiz: Decision Trees So Far

Week 3 Thursday September 19

BEFORE Class

1. Re-read S&Z pp. 216-219;
2. Problems TBA (Look ahead at problems for class below)
3. Re-read S&Z pp. 221-229
4. Problems TBA (Look ahead at problems for class below)

Class: Risk Aversion, Tree Flipping, and Imperfect Tests

• flipping trees
• risk
• Problems 102, 103, 104, 105, 106

**Lab 3: Decision Trees II - Information and Tree Flipping

• Instructions (TO BE REVISED)
• Deliverables TBA

Decisions Quiz

EXAM I AROUND THIS TIME

Tuesday September 24 - Thursday October 3

Difference Equations, Stock and Flow, and Feedback

+ Show Section Description

References and Resources

Connected Curriculum Project at Duke University. Difference Equations

If sequences and subscripts are at all new, watch this 8 minute lecture, "Sequences : Recurrence Relations : ExamSolutions : A-level Maths."

Sigma Notation
Not More Maths for Dummies 1.1: Sigma notation
Christine Breiner Summation Notation Practice (MIT 18.01SC Single Variable Calculus, Fall 2010)

Week 4 Tuesday September 24

BEFORE Class

1. If you are at all hesitant about using subscript notation, review it here.
2. Read Stokey and Zeckhauser, at least pp. 47-58 ch. 4 "Difference Equations"
3. Watch D Woodlock Introduction to Stock and Flow Diagrams
4. Ryan pages: 1 Introduction and 2 Rates and Amounts

• Problems TBA

Class Introduction to Difference Equations

• Problems 107, 108, 109, 110, 111

Thursday September 26

BEFORE Lab
Read Stokey & Zeckhauser pp. 66-73 and attempt problem 136

Class: Difference Equations II

• Problems TBA

Lab 5: Difference Equations

• Instructions
• Deliverable

Week 5 Tuesday October 1

BEFORE Class

1. Ryan Lecturettes: 3 Equilibria
2. Read S&Z pp. 66-73
3. Look over the table of contents in Kirkwood, System Dynamics Methods: A Quick Introduction] (DL)
4. Read Kirkwood, "System Behavior and Causal Loop Diagrams" (14pp) (DL)
5. Read Kirkwood, "A Modeling Approach" (6 pp. in DL)

• Problems TBA

Class: Stock and Flow Models I

• Problems TBA

Week 5 Thursday October 3

BEFORE Class

• Problems TBA

BEFORE Lab

• Problems 112, 113, 126, 131, 128, 129

**Class: Stock and Flow Models II]]

LAB 5: Stock and Flow Models

• Instructions
• Deliverable

Quiz Difference Equations and Stock and Flow Models

Linear Programing and Optimization

Week 6 Tuesday October 8

BEFORE Class

• If you feel at all shaking about graphing linear equations, avail yourself of tutorials on the net such as these at the University of Lancaster
• Problems  TBA

Class

• Getting Started
• Problems  TBA

Week 6 Thursday October 10

BEFORE Class

• Problems  TBA

BEFORE Lab

• Problems  TBA

Class

• Problems  TBA
• IN CLASS

LAB 6

• Instructions
• Deliverable

Part II Families of Models

W 10-Oct to T 16-Oct

Markov Models

+ Show Section Description

In this section (section 10 in Model Thinking), we cover Markov Processes. Markov Processes capture dynamic processes between a fixed set of states. For example, we will consider a process in which countries transition between democratic and dictatorial. To be a Markov Process, it must be possible to get from any one state to any other and the probabilities of moving between states must remain fixed over time. If those assumptions hold, then the process will have a unique equilibrium. In other words, history will not matter. Formally, this result is called the Markov Convergence Theorem. In addition to covering Markov Processes, we will also see how the basic framework can be used in other applications such as determining authorship of a text and the efficacy of a drug protocol.

Lectures and Reading

• Markov Processes
• 10.1: Markov Models (4:26)
• 10.2: A Simple Markov Model (11:27)
• 10.3: Markov Model of Democratization (8:21)
• 10.4: Markov Convergence Theorem (10:33)
• 10.5: Exapting the Markov Model (10:11)

Week 7 Tuesday October 15

Week 7 Thursday October 17

CHANGE

Linear Programing Lab 2

BEFORE Class

1. View lectures 10.1, "Markov Models" (4:26) and : 10.2, "A Simple Markov Model" (11:27).
2. Read S&Z 98-102 and "Markov Processes" pp 69-71
3. Problems 151, 157, 152, 154
4. Preliminary practice problem 151, 157, 152, 154

Week 8 Tuesday October 22

Class Introduction to Markov Models

BEFORE Class

1. View lectures 10.3, "Markov Model of Democratization," (8:21), 10.4, "Markov Convergence Theorem," (10:33), and 10.5, "Exapting the Markov Model," (10:11).
2. Read S&Z 102-114 and "Markov Processes" pp 72-82
3. Problems 153, 155, 156, 158
4. Do Markov problems in MT QUIZ 4.

Week 8 Thursday October 24

Class

LAB 6: Programing Markov Models

• Instructions
• Deliverable TBA

Section Quiz Markov Process Models

Actors, Others, and the Aggregation of Individual Decisions

+ Show Section Description

In this section of the course we contrast two types of models that explain a single phenomenon, namely that people tend to live and interact with people who look, think, and act like themselves, specifically, two well known models by Schelling and Granovetter.  The section finishes with what SEP calls "a fun model  about standing ovations" that he wrote with John Miller.  We will also see a computational version of Schelling's Segregation Model using NetLogo, free agent-modeling software created by Uri Wilensky or Northwestern University.

"Measuring Segregation"

BEFORE Class

1. SEP lectures 2.1 "Sorting and Peer Effects Introduction" and 2.2 "Schelling's Segregation Model".
2. Then look at and play with the NetLogo model online. Be sure to read (lightly) the documentation on the model's webpage.
3. Read either or both original chapter from Schelling's book and Page's notes on Schelling.
4. Finally, watch video lecture 2.3 on measuring segregation.
5. TBA
6. Quick review Quiz

Class: Schelling Segregation Model

• We will work on some follow-up exercises on this material.

BEFORE Class and Lab

1. View lecture 2.4 Peer Effects (6:57)
2. Read Granovetter Model
3. Read Schelling, Micromotives and Macrobehaviors, Chapter 3, pp. 83-133
4. View lecture 2.5 The Standing Ovation Model (18:04)
5. Read Miller and Page Model
6. Problems 165, 164, 170, 171, 173, 174, 175
7. Do Summary, overview, review questions.

Class

1. Peer Effects and Agent Models
2. More problems

LAB 8 : Title

• Instructions
• Deliverable

Sorting and Peer Models

EXAM AROUND THIS TIME

Tipping Points, Diffusion, and Contagion

+ Show Section Description

This section of the course is about tipping points. We focus on two models of "spread." The first is a percolation model from physics that we apply to banks and a model of the spread of diseases. The disease model is more complicated so I break that into two parts. The first part focuses on the diffusion. The second part adds recovery. The readings include two excerpts from the book SEP is writing on models, one on diffusion, the other on tipping (there is also technical paper by SEP and PJ Lamberson on tipping points included to provide a glimpse of what technical social science papers look like). Even if you don't read it in full, check out the introduction.

Week 9 Tuesday October 29

BEFORE Class
Read Tipping Points and Lamberson and Page: Tipping Points (READ INTRO ONLY)

1. View lecture 7.1, "Tipping Points" (5:58)

Class Tipping Points

Week 9 Thursday October 31

BEFORE Class

1. Read Diffusion and SIS
2. View lecture 7.2: Percolation Models (11.48)
3. View lecture 7.3A: Contagion Model 1-Diffusion (7:24)
4. View lecture 7.3B: Contagion Model 1-SIS (9:12)

BEFORE Lab

1. View lecture 7.4, "Classifying Tipping Points" (8:26), and lecture 7.5, "Measuring Tips for Measuring Tips" (13:39)

Class: Diffusion and Contagion

Lab 9 Tipping, Diffusion, and Contagion
Instructions
Deliverable

Section Quiz Tipping Point, Diffusion, and Contagion

Actors, Coordination, and Cooperation

+ Show Section Description

We will briefly look at various ways that social scientists model people, contrasting three different models:  the rational actor approach, behavioral models, and rule based models.  The first of these is the "model of man" (sic) that's implicit in much of our conventional policy thinking that applies micro-economics.  Then we'll have an introduction to "behavioral economics," a style of model thinking that takes into account the ways we know people tend to diverge from idealized rationality.  Lastly we look at rule based models.  Our take-away here is to be able to think about how behavioral considerations lead us to tweak the more straight-forward models based on rationality or rules. 

After this, we'll have a quick look at the question of cooperation and coordination. It may surprise you that this section will include a consideration of culture and game theory. It begins with a very simple game called the pure coordination game. In this game, a player wins only if they choose the same action as the other player. Which action doesn't matter as long as they choose the same one. The classic example is whether we drive on the left or the right side of the road. Either is fine as long as we all do the same. SEP will then consider situations in which people play multiple coordination games and study the emergence of culture.

Agent Models: Neighbors, Peers, Diffusion, Contagion

Week 10 November 5

Day 1

BEFORE Class we will assume you have viewed lectures on Schelling, Granovetter, and standing ovation model and read the associated materials.

IN CLASS we will develop a simple agent model in "pseudocode" and talk about the components of agent models in code.

HOMEWORK download NetLogo onto your work machine and work through tutorials 1, 2, and 3
Read Tipping Points and Lamberson and Page: Tipping Points (READ INTRO ONLY)

1. Read Diffusion and SIS

Day 2

AT SOME POINT view lecture 7.1, "Tipping Points" (5:58), 7.2: "Percolation Models" (11.48), 7.3A: "Contagion Model 1-Diffusion" (7:24), 7.3B: Contagion Model 1-SIS (9:12), 7.4, "Classifying Tipping Points" (8:26), and 7.5, "Measuring Tips for Measuring Tips" (13:39)

IN CLASS we will do some on-paper coding exercises on these models.

IN LAB we will play with a few NetLogo models and build one. We'll use either that or a pre-built one to collect some simulation data. Write up will be a short assessment of that data.

Week 10 Tuesday November 5

BEFORE Class

1. SEP Lecture 5.1: Thinking Electrons for Thinking Electrons: Modeling People (6:29)
2. SEP Lecture 5.2: Rational Actor Models for Rational Actor Models (16:09)
3. SEP Lecture 5.3: Behavioral Models for Behavioral Models (12:49)
4. SEP Lecture 5.4: Rule Based Models for Rule Based Models (12:30)
5. SEP Lecture 5.5: When Does Behavior Matter? (12:40)

**Class: Rationality, Rules, and Behavior

Week 10 Thursday November 7

**BEFORE Class and Lab

1. SEP Lecture 12.1: Coordination and Culture (3:37)
2. SEP Lecture 12.2: What is Culture and Why Do We Care? (15:43)
3. SEP Lecture 12.3: Pure Coordination Game (13:48)
4. SEP Lecture 12.4: Emergence of Culture (11:01)
5. SEP Lecture 12.5: Coordination and Consistency (17:03)
6. Coordination Games
7. Bednar et al. 2010
8. Axelrod Culture Model in Netlogo

Collective Action, Prisoners Dilemma, and the Commons

+ Show Section Description

Our next topic is collective action and "commons" problems with which you are probably familiar.

BEFORE Class

Start this section with the short introduction video (17.1) and then read the brief entry in the Stanford Encyclopedia and view lecture 17.2.

Complete this worksheet on prisoners' dilemma.

Read Nowak and Sigmund on cooperation and then view lecture 17.3.

1. SEP Lecture 17.1: The Prisoners' Dilemma and Collective Action (3:44)
2. SEP Lecture 17.2: The Prisoners' Dilemma Game (13:45)
3. SEP Lecture The Prisoners' Dilemma in the Stanford Encyclopedia of Philosophy
4. SEP Lecture 17.3: Seven Ways To Cooperation (15:20)

Read Nowak and Sigmund: "Five Ways to Cooperate"

1. SEP Lecture 17.4: Collective Action and Common Pool Resource Problems (7:23)
2. Read Ostrom: "Going Beyond Panaceas"
3. SEP Lecture 17.5: No Panacea (6:03)

Class

LAB 10

Section Quiz Coordination and Cooperation Models

Big Data and Visualization

Week 11 Tuesday November 12

BEFORE Class

Class

Week 11 Thursday November 14

BEFORE Class and Lab

Class

LAB 11

Week 12 Tuesday November 19

BEFORE Class

Class

Week 12 Thursday November 21

BEFORE Class and Lab

Class

LAB 12

Path Dependence and Mechanism Design

+ Show Section Description

Section 13 is about path dependence based on simple urn models. The take-away will be to unpack the logic of what makes a process path dependent and to see connections betwen path dependence and increasing returns and tipping points.

Mechanism design is about using insights like those we've talked about in this course to set up institutions that can encourage the production of pro-social outcomes. We begin with basics: overcoming problems of hidden action and hidden information and then turn to the more applied question of how to design auctions. The final lecture in the section looks at how to use mechanisms to aid in making decisions about public projects.
Cool Examples in D3
and others too, all by Jerome Cukier

Lectures and Readings

• SEP Path Dependence Quarterly Journal of Political Science.
• 13.1: Path Dependence (7:23)
• 13.2: Urn Models (16:26)
• 13.3: Mathematics on Urn Models (14:46)
• 13.4: Path Dependence and Chaos (11:08)
• 13.5: Path Dependence and Increasing Returns (12:31)
• 13.6: Path Dependent or Tipping Point (9:52)

Reading and Lectures

• 18.1: Mechanism Design (4:00)
• 18.2: Hidden Action and Hidden Information (9:53)
• Maskin: Mechanism Design
• 18.3: Auctions (19:59)
• V.S. Subrahmanian's auction slides
• 18.4: Public Projects (12:21)

Week 13 Tuesday November 26

To Do.

1. Start by viewing lectures 13.1 and 13.2 (about 25 minutes).
2. Next, work through a few practice problems which we will review in class.
3. Next read through SEP paper and view lectures 13.3 and 13.4 I(about 25 minutes).
4. Follow this with lectures 13.5 and 13.6 (about 20 minutes).

Week 14 Tuesday December 3

To Do.

1. View lectures 18.1 and 18.2. Work through these leading questions.
2. Read (intensely or not, as appropriate for you) Maskin. View lecture 18.3 and review the slides.
3. Finish up by viewing lecture 18.4 and working through these review questions.
4. Quiz 8 on sections 17-18.

Week 14 Thursday December 5

Models and the Wisdom of Crowds

The final section of the Model Thinking course looks at how diversity contributes to "wise crowds" when making predictions. It starts by reviewing category models and linear models and how they can be used to make predictions and then introduces the Diversity Prediction Theorem. The course concludes with a lecture on the value of having lots of models.

• Diversity Prediction Theorem
• 20.1: Prediction (2:25)
• 20.2: Linear Models, Categorical Models, and Prediction (5:02)
• 20.3: Diversity Prediction Theorem (11:54)
• 20.4: The Many Model Thinker (7:11)

[[/collapsible]]

Section 20: The Many Model Thinker: Diversity and Prediction]]

To Do.

1. Listen to lectures 20.1 and 20.2 (about 7 minutes) and then read Page on the "diversity prediction theorem."
2. Listen to lectures 20.3 and 20.4
3. Work through these problems/questions.
4. Do Quiz 9 but skip the material no chapter 19.

Thursday 5 December

Class: Review
Lab: Take-Away Skill Show

FINAL EXAM Mon Dec 16 2–5pm