Syllabus

Why are we here?

Data is one powerful way to learn about how the world is and how the world works. Quantitative arguments are pervasive, often intersecting with moral or philosophical issues in international relations, domestic politics, personal lives, and many other domains. The overarching goal of the course is to develop your ability to make and assess quantitative arguments. In particular, you will practice:¹

• translating important questions into quantitative frameworks,
• selecting and conducting appropriate methods of quantitative analysis,
• drawing substantive insights from your quantitative analysis, and
• communicating your methods and conclusions so that others can understand and interrogate your quantitative reasoning

What will we learn?

The course is organized into five units that will develop the capabilities above. The first four will equip you to address essential questions of quantitative reasoning:

1. Descriptive statistics: How can we synthesize large quantities of information to establish general, actionable facts about the word? E.g., based on surveys of thousands of Rohingya refugee households in Bangladesh, what do we know about mental health in this community? 2 Regression: What is the nature of relationships between key measures of environments, behavior, or well-being? E.g., how are traumatic experiences related to mental health concerns among refugees?
2. Inference: Do these facts or relationships generalize? E.g. what can we say about mental health, and its relationship to traumatic experiences, in the larger population of almost one million refugees?
3. Causality: How can we identify causal relationships in data? E.g. what programs lead to improvements in psychological well-being?

In the final unit, you will, as a part of group, produce an original research project demonstrating your ability to make a quantitative argument.

In parallel to statistical concepts, you will learn the computing tools required to actually conduct quantitative analysis. This course uses the software R because it is free and open-source, powerful but approachable, and a job-relevant skill.

Course resources and website

Our primary guides (and required texts) will be

• Naked Statistics: Stripping the Dread from the Data,
• Thinking Clearly with Data, and
• the SIS 600 R Survival Guide.

Naked Statistics and Thinking Clearly with Data are available at the University library. There are also affordable paperback versions in case you would like to purchase them. The Survival Guide is a free online resource maintained by the SIS 600 instructors.

You will discover that these resources overlap. Periodically, you will be assigned to read sections from different texts on the same topic. That may seem redundant, but, faced with challenging material, I find students learn better by seeing similar content presented in different ways.

Canvas hosts the course website, where you can find all other course materials, including assignments, weekly meeting materials, and extra resources for learning R.

How will I be evaluated?

Assignments

Regular practice is key to consolidating the type of skills that are the focus of this class, so you will have a regular cadence of low-stakes quizzes and homework assignments. You will also take an in-class midterm exam, deliver an in-class presentation, and complete a group research project. The assignments include a variety of ways for you to demonstrate your understanding, including structured problems, interpretive questions, oral presentations, and original written work.

The assignments and their contributions to your final grade are summarized below. Due dates are listed on the course outline. You can also find due dates on Canvas, alongside complete instructions for each assignment. Canvas is also the place where you will submit all assignments except for the quizzes.

Quizzes (10 percent)

Most weeks, we will start class with a 15-minute, pen-and-paper quiz intended as a low-stakes incentive to prepare for class and to consistently engage with course material. The first quiz will be a leniently graded and cover basic information from the syllabus. The remainder will be based on the reading assigned for each class meeting, offering another way for me to highlight key takeaways from the reading.

There are a total of nine quizzes. Your lowest two scores will be dropped, no questions asked, so the remaining seven in-class exercises will each be worth a bit less than 1.5 percent of your final grade.

Homework assignments (30 percent)

Roughly every other week, I will assign multi-part computer-based homeworks to be submitted via Canvas. The first homework assignment focuses on basic math topics relevant to the class. The remainder will reinforce material covered in the preceding class.

There are a total of seven homework assignments. Your lowest two scores will be dropped, no questions asked, so the remaining five assignments will each be worth 6 percent of your final grade.

Midterm exam (20 percent)

A computer-based midterm exam will be administered in class during our Week 9 meeting. The exam will be very similar in content and format to the homework assignments.

Research project (40 percent)

A primary goal of this class is to enable you to conduct original quantitative analysis on a question that motivates you. That goal motivates a research project with three stages:

• Individual project planning worksheet (5 percent)
• Individual presentation (10 percent)
• Group final paper (25 percent)

Complete details of the research project are available on Canvas.

Assignment grades

For quizzes, homework, and exams, each question (or, in the case of a multi-part question, each part) of an assignment will be worth one point. Each question (or part) will be scored according to the following rubric:

• 1 point: a full and correct answer
• 0.5 points: any reasonable attempt, i.e. some evidence of effort
• 0 points: no reasonable attempt, i.e. no evidence of effort

Grappling with difficult content is a part of the learning process. Even if you are stumped by a question, you will be rewarded for trying.

For two reasons, I strongly encourage you to review all graded assignments. First, comparing your submissions to the solutions is an effective way to learn. Second, I make mistakes sometimes. You have two weeks from the time an assignment grade is posted or revised to raise concerns. After two weeks, you have accepted the grade and I will not revisit it even if we later determine that I made an error.

Course grades

Course grades are curved. That means the standard correspondence between percentage scores and letter grades (i.e. 90-100 percent earns an A, 80-90 percent earns a B, and so on) may not apply in this class. As such, I urge you not to focus on the percentage scores. Instead, I will provide regular updates to your cumulative letter grades throughout the semester via Canvas.

Monitor your cumulative letter grade closely. The MA programs require a C or better in this class to graduate.

Late work

A policy regarding late work must balance competing concerns. One concern is empathy; life is complicated and unexpected challenges constantly arise. Another is equity; while one student turns in work late, another may make an incomplete submission on time. Some students feel more comfortable than others requesting extensions. And I am obliged to delay posting solutions to assignments until I have received all submissions.

Late assignments submitted within 48 hours of the Canvas deadline will receive half of the points they would have earned if they had been submitted on time. Late assignments submitted 48 hours or more after the Canvas deadline will receive zero points.

There are no make-up assignments. To account for mild illness, competing work deadlines, family commitments, etc., your two lowest quiz scores and your two lowest homework scores will be dropped. If you are facing extenuating circumstances, please contact me as soon as possible.

Group work

Working with classmates often improves learning and performance on assignments. I encourage you to collaborate.

However, with the exception of the research project final paper, the written work you submit must be yours and yours alone. You may discuss concepts, work through programming errors, and compare results as a part of a study group. But you are expected to make original, individual submissions for each assignment. You may not send electronic files or copy and paste material written by another person. If you have any questions about what constitutes an original, individual submission, please ask.

The research project final paper is a group assignment. We will form groups of three to five based on your interests, and each group will make a single submission.

On AI tools like ChatGPT

AI tools like ChatGPT are advancing remarkably fast. We are all still learning how to use these new technologies, including their place in education. My current view is that AI tools can help you learn, but they should be approached with care. Here are a few policies and principles:

• You may use ChatGPT and similar tools to assist you in completing at-home assignments, but make a good-faith attempt at each problem before turning to an AI tool.
• Anytime you use an AI tool to assist you with an assignment, please submit the inputs and outputs from the tool as an appendix to the assignment.
• Outputs from AI tools can be incorrect, incomplete, biased, and / or improperly cited. You are responsible for the quality and integrity of your own work.

What will a class meeting be like?

Most classes will start with a 15-minute, pen-and-paper quiz based on the reading for that day. We will review the quiz and proceed with instructor-led session covering that meeting’s content. Class will usually conclude with an R-based application and a preview of the following week’s material. You can count on a 10-15 minute intermission around mid-way through each meeting.

What if I don’t feel comfortable with math or statistics?

First, know that you are not alone. Many people feel daunted by math and statistics including, in my experience, most students in SIS 600. I believe math is hard for [almost] everyone.² The trick is struggling patiently and not getting discouraged. Of course that’s easier said than done, especially for the many people who have been convinced over a lifetime, incorrectly in my view, that they are “bad at math.”³ So I strongly urge you to ask for help and encouragement when you need it. See the section below “Where can I go for help?” for suggestions.

The course assumes no prior background in statistics, statistical computing, or advanced math. What you will need is working knowledge of several basic mathematical concepts, including order of operations, arithmetic, percentages, slopes, summation notation, and basic algebra.

To make sure you these concepts are fresh in your mind, you will find a variety of tutorials, instructional videos, and practice problems here. The syllabus and math quiz will provide you with an opportunity to practice and demonstrate relevant skills.

How can I do well in this class?

Standard guidance applies; you will benefit from coming prepared, participating actively, etc. But there are a few recommendations that, in my view, are particularly important for this course.

Be consistent. Put in the time every week. Regular engagement with the material and showing up for class prepared are especially valuable for a subject like statistics. Building familiarity with the R software will pay dividends as you develop your research project.

Ask for help. Office hours are my favorite part of teaching. I enjoy getting to know students a little better and small-group interactions are an effective, efficient way to deepen your understanding. Please come see me. I also encourage you to visit your TA, who not only has a strong grounding in the content but also the perspective of a successful graduate of the class.

If needed, do more. Knowing that you face many competing demands, I am selective about what I require for the course. But sometimes the minimum requirements may not be enough for you to completely master a new concept or skill. Some supplemental or practice materials will be available on Canvas. One of our books, Thinking Clearly with Data, includes exercises at the end of each chapter.

Where can I go for help?

• Canvas and syllabus should be your first stop for administrative questions.
• Our primary guides Thinking Clearly with Data, Naked Statistics, and the SIS 600 R Survival Guide, combined lecture notes, assignments, and solutions on Canvas, are your curated sources for course content.
• There is no shame in searching online! Most of what I know about statistical computing I learned from Google searches. Just be cautious. Online resources can be confusing, e.g. because they use unfamiliar mathematical notation or R syntax.
• You may use AI tools like ChatGPT subject to the policies described above.
• There are many humans who will help you. Your classmates are a great resource; I survived graduate school by working with groups. I invite you to be in touch with your TA or me, your professor.
• For outside-the-classroom difficulties, please contact the Dean of Students Office and know that the university offers counseling services and victim-survivor advocacy services. If you feel comfortable doing so, please also reach out to me, your professor. The Academic Support and Access Center supports students with disabilities. I want to enable full participation of all students. Please reach out to me as soon as possible to arrange any necessary accommodations.

Contact and office hours

You can reach me via email at cadavis@american.edu. My office is SIS 347. I have office hours:

• Monday 10-11am (in person OR Zoom)
• Tuesday 2-3pm (in person OR Zoom)
• Friday 4-5pm (Zoom only)

Please let me know you are coming by making an appointment here.

Your TAs will also be hosting office hours:

• Pietro Zecca - Tuesday 4-5pm - SIS 120
• Gustavo Maragoni - Wednesday 4-5pm - SIS B60

SIS B60 is located inside the Graduate Research Center (GRC) on the Terrace level. The GRC is access controlled; you will need to swipe your AU ID for entry. If you do not already have permission, you can request it by emailing access@american.edu.

Academic integrity and other University policies

I expect you to observe relevant university policies, including the Academic Integrity Code and health-related precautions. The University maintains a website with the latest guidance. Unauthorized recordings of classroom meetings or class-related content are prohibited.

Course outline

Descriptive statistics

Week 01 (Mon Jan 12 | Tue Jan 13)

Topics

• Orientation
• What is data
• Introduction to R

Reading

• Syllabus
• Thinking Clearly with Data (TCwD) Ch 1 (on Canvas)
• Naked Statistics (NS) Ch 1 (on Canvas)
• R Survival Guide Ch 1-2

Assignments

• Install R & R Studio

Week 02 (Mon Jan 26 | Tue Jan 20)

Topics

• Visualizing distributions
• Comparisons across groups

Reading

• NS Ch 2
• R Survival Guide Ch 3-4, 6

Assignments

• Homework (math assessment)
• Quiz

Week 03 (Mon Feb 02 | Tue Jan 27)

Topics

• Central tendency and spread
• Describing trends

Reading

• NS Ch 3
• R Survival Guide Ch 5, 9

Assignments

• Homework
• Quiz

Regression and relationships between variables

Week 04 (Mon Feb 09 | Tue Feb 03)

Topics

• Visualizing and measuring associations between variables

Reading

• NS Ch 4
• TCwD Ch 4

Assignments

• Homework
• Quiz

Week 05 (Mon Feb 16 | Tue Feb 10)

Topics

• Regression with one explanatory variable

Reading

• NS Ch 11
• TCwD Ch 5

Assignments

• Homework
• Quiz

Week 06 (Mon Feb 23 | Tue Feb 17)

Topics

• Regression with several explanatory variables

Reading

• TCwD Ch 10

Assignments

• Homework
• Quiz

Inference

Week 07 (Mon Mar 02 | Tue Feb 24)

Topics

• Probability
• Central Limit Theorem

Reading

• NS Ch 5, 8

Assignments

• Research project planning worksheet
• Quiz

Week 08 (Mon Mar 16 | Tue Mar 03)

Topics

• Sampling distributions
• Standard errors

Reading

• NS Ch 9
• TCwD Ch 6

Assignments

• Homework
• Quiz

Midterm exam

Week 09 (Mon Mar 23 | Tue Mar 17)

Topics

• None

Reading

• None

Assignments

• Midterm exam

Causality

Week 10 (Mon Mar 30 | Tue Mar 24)

Topics

• Common mistakes
• Defining the counterfactual
• Randomized controlled trials

Reading

• NS Ch 12, 13
• TCwD Ch 9, 11

Assignments

• Quiz

Week 11 (Mon Apr 06 | Tue Mar 31)

Topics

• Diff-in-diff
• Regression discontinuity
• Instrumental variables

Reading

• TCwD Ch 12, 13

Assignments

• Homework
• Quiz

Research project

Week 12 (Mon Apr 13 | Tue Apr 07)

Topics

• None

Reading

• None

Assignments

• Research project presentation

Week 13 (Mon Apr 20 | Tue Apr 14)

Topics

• Advanced R: merging and data cleaning

Reading

• None

Assignments

• None

Week 14 (Mon Apr 27 | Tue Apr 21)

Topics

• Advanced statistical techniques
• Project workshop

Reading

• None

Assignments

• None

Footnotes

1. These learning outcomes are lightly adapted from AU’s Quantitative Literacy (Q2) curriculum.

2. There is an admittedly niche sub-genre of quotations in which genius mathematicians humblebrag about how deeply mysterious mathematics is. For example, John von Neumann reportedly said “In mathematics, you never understand things; you just get used to them.” According to Bertrand Russell, “mathematics is the only subject in which we never know what we are talking about or whether what we are saying is true.” As a small reward for carefully reading the syllabus, you will receive one percentage point of extra credit for emailing me a one-sentence description of a major contribution that von Neumann made to my field of economics. You may submit this extra credit at any time before the final.

3. For example, the unfounded stereotype that women are “bad at math” appears to impair performance. Some relevant evidence is summarized and cited in this article from Science.