A typical week will look something like this:
Each of you will attend one synchronous meeting per week via Zoom. Please only attend the meeting you signed up for on Banner.
I expect you to watch the relevant videos before our synchronous meetings. The “relevant” videos are the ones scheduled before the meeting. After you have watched the videos, you can complete short quizzes to check your understanding. You will get class credit for completing these.
This course is about asking and answering questions with data.
A lot of people think about research and statistical methods as something other people do. The goal of this course is for you to develop the capability to understand research other people have done, and to conduct independent research of your own. I want you to develop the confidence and knowledge to ask important questions and to generate scientific evidence. I strongly believe that regardless of your background, by the end of this course, you will be able to do quantitative research.
This course is also about building up your ability to read and write quantitatively. You will likely struggle at times. This is good! We often learn the most when we struggle a bit. Don’t shy away from trying, even if it doesn’t work out right away. Think of me as your statistics coach. I’ll challenge you a bit to help you learn and develop, but fundamentally my goal is to help you get the most out of this course. Don’t hesitate to ask for help.
By the end of this course, you will:
Be able to examine and describe data
Understand the connection between sample statistics and random variables
Be able to create compelling visualizations to describe data and convey insights
Understand the difference between correlation and causation
Appreciate the central limit theorem (the most important theorem in this class!)
Write models to frame a question
Estimate a multivariable linear regression
Test a hypothesis
Conduct empirical research
Use two statistical programming languages (R and Stata)
The major theme is statistics as a lens for understanding inequality. We will cover statistical content including descriptive analysis, visualization, hypothesis testing, inferential errors, and regression modeling. We will learn these topics by studying important and pressing questions relating to inequality, including:
Mass incarceration
Racial income gaps
Coronavirus case rates and income distributions
Income inequality and public sentiment
Climate change and public sentiment
Air pollution and environmental justice
Statistics offers one lens through which to view these topics. While it is a powerful lens, it is not the only one. The conclusions we can draw using statistics rely on our engagement with subject matter experts, particularly non-statistical accounts reflecting lived experiences. In econ-speak, quantitative analysis is a complement (not a substitute) to qualitative work. In addition to a useful statistical toolkit and the knowledge and confidence to apply it, I want you to take from this class a sense of humility about its limitations when used in isolation.
I identify (among others) as an Indian-American man, an economist, a teacher, and a writer. I speak and write primarily in English and have spent my life financially comfortable. Most of my life has been in India and the US. My class, gender presentation, degrees, intellectual pursuits, and American accent are often more prominent in the spaces I inhabit than my ethnic identity, family background, or life experiences.
While issues of global environmental justice are very close to me personally and intellectually, I have no lived experience and little prior scholarly engagement with topics relating to Black experiences in the US. I hope to learn more about these experiences and discuss them with respect and sensitivity.
The topics and examples in this class reflect, among other things, my desire to engage more with issues of inequality and social justice. I recognize that some of these topics may have deep emotional resonances. I care deeply about not perpetrating harms by discussing issues inappropriately. While it is not your responsibility to educate me when I slip, I hope we can develop the mutual trust to support calling each other into more inclusive and empowered discussions. The topics and examples in this class also reflect my academic US-centric perspective. I recognize that many of you bring other perspectives to our discussions. I aim to support you if and when you choose to express your different perspectives.
I am committed to making this learning experience as fruitful as possible for you. Over my life I have struggled a lot with math, but have been exposed to computers and programming from an early age. I recognize that my personal background shapes what I do and don’t immediately recognize as challenging. I hope you feel comfortable asking me to explain in more detail if something is confusing.
At Middlebury, we strive to make our campus a respectful, engaged community that embraces difference with all the complexity and individuality each person brings. Each student in this course is expected to contribute to an inclusive and respectful class environment. Students of all backgrounds including gender, sexual orientation, race, ethnicity, and religion are to be treated fairly and with honesty, integrity, and respect. Civil discourse without degrading, abusing, harassing, or silencing others is required of all students in this class.
Being intentional with our words and actions is always important, and particularly so when we discuss heavy topics like those in this class. Please, be mindful of how you frame points and arguments. Avoid pathologizing language in particular—it is often both harmful and a marker of a poorly-considered argument. Intentional and thoughtful discussions which do not perpetrate harms increase our collective freedom of speech and make it possible for us all to engage more fully. Similarly, when someone notifies us that we have perpetrated harms, robust intellectual discourse is best served when we listen, apologize, reflect, and adjust our behaviors moving forward.
We’ll be using Introduction to Statistics with Randomization and Simulation (ISRS). The book and all associated materials (lab materials, code, videos) are free and open source; just go to the link and download/stream what you want. If you want a paper copy, you can pay a low fee for printing and shipping.
We’ll follow the textbook somewhat closely, at least until the later chapters. In later chapters we’ll start to deviate a bit more to focus on issues more relevant to economics and social science. You can safely ignore/skim any statistical concepts/topics we don’t discuss in videos/labs/deliverables. When in doubt if you should pay attention to something in the textbook or not, don’t hesistate to send a Slack message asking for clarification (preferably a public message in the general channel). In general I expect you to have skimmed the relevant textbook chapters prior to our meetings.
Sometimes class notes will draw on other sources; these will be referenced, but you are not expected to look at those references.
We will also read academic articles, essays, and book excerpts to complement our understanding of the substantive topics in the course. I will make these readings available on Canvas.
We’ll be using R and RStudio for the majority of this class. R is a free open-source statistical software package, and RStudio is a popular integrated development environment which makes it easier to work in R. As we get into regression modeling near the end of the class, we’ll use Stata. Stata is not free and open source, but we’ll have free access via Apporto. If you are experiencing connectivity issues using Apporto and Stata, let me know and we’ll figure something out.
In general if your question is about specific code or math, please include a copy of the code (as a .Rmd file or code chunk) or the math (as a PDF, picture, screenshot, or text). I (try to) maintain work-life balance by not checking messages after 5pm Eastern time or during the weekends, but I promise that I will respond to any message you send me. If it’s been more than 48 hours and you haven’t heard from me, please follow up.
My Zoom room ID is 927 138 2648. All of our class meetings will be held here.
Think of Zoom as the equivalent of a classroom (during scheduled meeting times) or my office (all other times)—it’s rude to interrupt someone else’s scheduled time. Please use Slack or email if you want to ask a question not during scheduled office hours or appointments.
I’ll be in my Zoom room during scheduled meeting times and office hours. Outside of those times, please contact me through Slack or email.
Use this invite link to join our class workspace (it expires after 30 days—let me know if you need an invite). If that link doesn’t work, you can also use this link to sign up for the workspace (this one requires you use your @middlebury.edu email). You can access the workspace using your browser at this link, or from the app.
I’ll be on Slack more frequently (basically whenever I’m at my desk but not shutting off all distractions to work). Use this for questions that don’t need a longer reply or discussions with each other. Please use the appropriate channel for your questions/discussions: problem set-specific questions should be in problem-sets, general course logistics questions should be in general, lab questions should be in labs, etc. Anything that doesn’t have a specific channel should go into random or general.
If you are asking me a question on Slack or would like me to weigh in on a discussion, please tag me (type @ProfRao once somewhere in your question). I’ll check Slack at least once per day during the week. I won’t be checking Slack over the weekend. If you ask me a question there and I don’t get to it right away, I’ll tag you in my reply. You can also DM me with questions if you don’t want to post them in a public channel. I may post such questions (anonymized) to a public channel if I think it would help others in the class.
Think of Slack as an in-between space where you can ask quick questions and have unscheduled discussions—not as structured as Zoom (also, no video), but more fluid than email.
Email is useful for longer questions, and allows more detailed responses. In the subject line, please start with “Econ 210:”. I use email filters to prioritize among the many emails I receive each day; not including the “Econ 210:” subject line will delay my response. Please allow me up to 48 hours to respond to emails. Don’t hesitate to follow up if I have not replied (you can just re-send the email).
If your question is about specific code or math, please include a copy of the code (as a .Rmd file) or the math (as a PDF, picture, screenshot, or text description in the email).
The office hours setup this semester is a bit different from how I usually do it. We’ll have two types of sessions: problem-solving sessions dedicated to specific problems submitted in advance, and drop-in sessions dedicated to whatever questions you ask.
These sessions will be dedicated to going over pre-submitted questions. You’ll submit any questions you want covered in these sessions to the office-hours channel. I’ll keep a running list of questions that haven’t been addressed yet, and we’ll work through them during the session. I’m open to recording these and posting on Panopto/Canvas if the times prove challenging—just let me know if it would be helpful. I’ll mention in the office-hours channel if we’re recording the current session.
If we run out of pre-submitted questions, we’ll use the remaining time as a drop-in session.
These are the most like normal office hours. Questions will be answered round-robin in order of entry to the Zoom. You can also submit questions on Slack in the office-hours channel if your Zoom isn’t working.
I love meeting with students. But! Please make an appointment if you want to meet outside of office hours. A quick email or Slack DM is fine. Friday is my “research day”, so I try not to schedule appointments then, and may be slow to respond.
This course may seem daunting—we’ll be covering complicated concepts, sometimes at a very abstract level. Don’t panic! With practice, repetition, and patience, it will come together. I’m here to help—don’t hesitate to ask questions.
On asking questions: sometimes a concept feels fuzzy or like it isn’t quite clicking, but it’s hard to frame a question precisely. Don’t worry! Ask anyway. We may not get to the bottom of it right away, but we can make progress. Questions about material also tend to be correlated across students, so by asking (especially in a public Slack channel) you often provide a public good to your classmates.
This material goes deep. I may at times limit our inquiry to keep us moving along. If you’re curious, I’m happy to talk in more detail in office hours. Think of statistics as a very big onion, and this course as pulling the first few layers off—you can chop those layers up and cook with them, plenty of recipes won’t need any more!—but there’s a lot of onion left.
Statistical programming proficiency is an important outcome of this class. I strongly encourage you to work through the code examples in the textbook and lab notes (actually type the commands in, don’t just copy-paste). You can look at the lab notes attached to the textbook for more examples. Stack Overflow is a particularly helpful online resource for programming questions in R and other languages. I encourage you to spend some time getting comfortable with using the site.
There are a number of deliverables in this class. You do not have to do all of them to get an A (or whatever grade you’re after). See the grading contract below for details on what you need to do to achieve specific grades.
Video responses: Professor Gong and I will post videos covering statistical and programming concepts. After you watch a video, you can take a short multiple-choice quiz testing recall and understanding.
Lab reports: I will provide you with lab notes prior to each lab. These notes will contain exercises for you to do. The lab report for a week is an Rmarkdown file with your work on completing the exercises.
Problem sets: These are weekly collections of problems to solve. A completed submission is an Rmarkdown file with your work towards solutions. When you are unable to completely solve a question, sketch out your work so far, your thoughts on what the answer ought to look like, and what steps you think need to be taken to get there.
Problem set corrections: When you submit your problem set, you will receive access to the solutions. You can review the solutions and submit a “redo” file explaining what you got wrong. If you got everything right, you can just say so.
Challenge questions: These are more challenging/involved questions intended to test your understanding and stretch your skills. You will have two weeks per challenge question.
UpShot-style blog post: These are the capstone research project for this class, blending statistical analysis and communication.
Please feel free to ask me questions about any assignments. However, you will probably get the most out of questions if you ask them after working on them a bit and running into an issue.
After each video, you can take a short multiple-choice quiz to test your recall and understanding of the video you just watched. The quizzes are accessible on Canvas.
Video quizzes have a small point value attached to reward the effort of watching the video.
The lab report notes will guide you through the mechanics of the new code we’re learning that week. The exercises are designed to give you some practice and (sometimes) lead you to new insights. You will automatically get credit for submitting a completed lab report whether the answers are correct or not. You’ll submit lab reports on Canvas.
I recommend not spending more than 2 hours on a single lab report. There are diminishing marginal returns to staring at the same thing. Eventually we all hit a wall of “I’m clearly not getting something but I don’t know what and it’s not getting any clearer.” You won’t be penalized for wrong answers, so you might as well submit when you hit the wall.
I want to incentivize you to put effort into the problem sets even if you aren’t sure you have the right answer. My criteria for “effort” is “the question was attempted and completed, or attempted and incomplete with a brief explanation of where and why the student got stuck”. “Brief” means “enough to convey understanding and the issue”, not “everything the student knows about the subject”.
After you submit, you will get access to the solutions. You can then submit problem set corrections: revise your solutions to each question you got wrong and resubmit for credit. If you got the question right, just say so. This is to give you an incentive to review the solutions, even if you got everything right. The idea here is that the effort of reviewing solutions is rewarded, whether you got the answer right or not.
I recommend not spending more than 3 hours on completing a single problem set.
One of my favorite things about the toughest questions on well-designed exams is that they teach me something new. We don’t have exams, but I’d still like to give you some tougher questions that stretch your skills a bit more and maybe teach you something new and useful. The challenge questions fill that role. Every two weeks, I’ll post a new challenge question. You can discuss the questions and the issues you face with others, but I think you’ll get the most out of these questions if you first make a serious attempt on your own. They’ll often deal with real issues researchers study; if so, I’ll post some underlying research the question is based on after the challenge question is due, and either we’ll discuss it a bit in class or I’ll make a video (or both, I’m not sure yet). You’ll submit challenge questions on Canvas as Rmarkdown files detailing your work.
I recommend not spending more than 4 hours on a single challenge question.
In previous versions of this class, I’ve assigned a research project. This was typically a two-person project where students produced an academic-style paper showing off their statistical and programming skills on a question they found interesting. I’ve come to think this isn’t the best assignment for a couple reasons. First, though free riding was rare, each instance led to significant discomfort for those picking up the slack, less learning for those exerting less effort, and a lower-quality product for me to assess. Second, the academic writing style for statistical research will probably not be relevant to most of you later on, and students spent a lot of energy mimicking that particular writing style. Nobody who hasn’t spent years stewing in academia likes to read a typical academic research paper dealing with statistics; even the best-written ones generally have low readership.
The UpShot, on the other hand, is widely read by academics and non-academics. Each post is generally well-written, has an interesting question and a key point, and gets to the point fast. It features statistical analysis and visualizations as needed to drive the key point home and address likely reader questions. A non-specialist could read an UpShot post and come away smarter. Whatever you do later on, it’s (probably) the kind of output you’ll need or want to produce to convey quantitative insights to a broad audience.
In the honor code section I say you can collaborate on anything “except the blog post”, which is worth clarifying. You are free to discuss your blog post with others, to review each others work and code, and even solicit suggestions. But you must submit your own work, reflecting your own effort and interests. It’s fine if you end up asking the same question as someone else in the class (this happens surprisingly frequently in academic research), but you must have your own analysis and interpretation. This doesn’t mean you need to disagree with others working in similar areas.
Where you use someone else’s ideas, you must cite them. This applies to any source you use: popular press articles, academic sources, etc. Use inline citations (not footnotes) in the Chicago Author-Date style used in Economics or hyperlinks. Your bibliography should sit between the main text and statistical appendix.
You’ll submit the blog posts as html or PDF files (your choice) on Canvas.
I recommend starting the blog post early.
The effort tiers here are cumulative, i.e. to get a “complete” you must satisfy the “basic” and “complete” requirements. I will assess whether a post meets the criteria for a given tier or not with feedback on suggested improvements. If you are unsatisfied with my assessment, you can redo the blog post up until the final posting date for blog posts (December 11, 2020). There is no limit on the number of redos. This assignment is inspired by Professor Wolcott, and many of the criteria below are borrowed from her rubric. Blog posts at any level must include a bibliography. I will provide an assessment rubric around week 5 to help guide your writing.
Grading is problematic. It induces anxiety and stress, fosters competition amongst students, induces proclivities towards honor code violations, and focuses student time and attention on grades instead of learning. Many professors also think it is the worst part of our jobs. Grading takes up a lot of time and typically doesn’t result in timely or productive feedback. Finally, grading can create an adversarial relationship between professors and students; office hours discussions around grades (e.g. “What can I do to get a better grade?”, “Can I do extra credit?”) are generally unpleasant for all involved.
All that said, I still have to assign you a final grade in the course.
Instead of receiving a final letter grade based on mastery of course materials, I will reward you based on the effort that you put into the class. In a traditional statistics course, students take exams and receive grades based on the number of correct responses (grades are based on knowledge). In this course, students will be given assignments, and receive credit based on the effort they put into these assignments (grades based on labor).
The underlying theory here is that exerting effort has benefits even when it feels like you’re not getting much out of it. I want to rewire the usual effort-output-reward loop a bit so that you’re more directly incentivized to do things that will (on average) help you learn, even if in the moment it feels like a struggle and you’re not getting anywhere. From what I’ve seen, all else equal,
I start from the premise that if you’re in this class, you want to (a) learn statistics and statistical programming and (b) achieve a certain grade. All too often, course (and other) incentives students guide students to focus on more on getting good grades than on actually learning. With this system, I hope to remove uncertainty from your final grade (you can achieve the grade you want with a relatively well-understood effort allocation) and make it easier for you to learn (less anxiety about assessments and more focus on process). Think about this question: “If I just wanted to learn this material and wasn’t getting a grade for this assignment, would I still do it?” If the answer is “No”, then maybe the assignment isn’t the best use of your time.
I want to take the emphasis away from grades, and put it back on where it belongs: the process of learning.
Finally, labor-based contracts create a more (but not totally) equitable environment. Using a traditional grading system, students that have prior knowledge of the topic have a distinct advantage. With labor-based contracts, all students must put in time and effort; you cannot coast based on what you already know.
There will be a set of assignments that you can choose to do. When you submit an assignment, it is either accepted (receive credit) or declined (no credit). Anything that requires effort will be eligible for credit. Since we are focused on your effort and not knowledge, there will be no exams in this course. (You heard that correctly. No exams. A godsend to both students and professors.)
Each grade requires you to obtain a number of points in the course. You can get points by completing assignments and receiving credit for them. For the UpShot-style blog posts only, I will track variation in output quality. The tiers of blog post quality are discussed in the “Blog posts” section of “More details on deliverables”. Every assignment except the blog posts will be assessed on an accept/decline basis.
The table below shows one (recommended) approach to securing specific grades. The point requirements for each grade are listed alongside the recommended quantity of assignments to complete to get there. The points attached to a marginal assignment of each type are in the subsequent table.
| Grade | Points required | Video quizzes | Lab reports | Problem sets+corrections | Challenge questions | Blog posts |
|---|---|---|---|---|---|---|
| B | 91 | 19 | 6 | 6 | 0 | 0 |
| B+ | 139 | 23 | 7 | 7 | 1 | 1 basic |
| A- | 175 | 23 | 8 | 8 | 2 | 1 complete |
| A | 217 | 25 | 9 | 9 | 3 | 1 extensive |
| Assignment | Point value |
|---|---|
| Video quizzes | 1 |
| Lab reports | 6 |
| Problem sets | 3 |
| Problem set corrections | 3 |
| Challenge questions | 12 |
| Blog post (basic) | 20 |
| Blog post (complete) | 32 |
| Blog post (extensive) | 48 |
The point values are meant to be loosely proportional to how helpful I think each assignment is, at the margin, in furthering your learning. Credit is assigned on an accept/decline basis for all assignments except the blog post. They also provide a convenient “exchange rate” to use between assignments, so that you have options in case you’re unable to meet the recommended requirements for the grade you want.
Definitely! No, just kidding. Believe it not professors are really busy (just like students). And there is an opportunity cost for everything that we do. Instead of spending hours marking off points on a student’s exam, we want to spend our time providing students with feedback on what matters: their research projects and learning.
Yes! True understanding of something is a result of the effort you put in, not a 24 hour cramming session before an exam. In fact, our (informal) best guess is that students forget most crammed things within a week after an exam.
The beauty of this system is that you get credit if you put effort into answering the question, and not the actual answer. I will challenge your thinking with questions that require a lot of effort. I expect you to put in the effort. However, I want to minimize any anxiety about asking you tough questions because you don’t have to worry about getting the correct answer. As long as you show a good-faith effort, you’ll get credit.
Great question. I hope so.
The college has many resources to help you when you are struggling. You can find a list of resources here.
I’d like to make a particular mention of the Anderson Freeman Resource Center. The AFC has some great programming, including a peer mentorship program, peer writing tutoring, wellness support, counseling, and social events (granted we’re in a pandemic, but even online events can be nice). The AFC is physically located at Carr Hall and can be reached at afc@middlebury.edu or 802-443-2214.
Students who have Letters of Accommodation in this class are encouraged to contact me as early in the semester as possible to ensure that such accommodations are implemented in a timely fashion. For those without Letters of Accommodation, assistance is available to eligible students through Student Accessibility Services. Please contact Jodi Litchfield or Peter Ploegman, the ADA Coordinators, for more information: Peter Ploegman can be reached at pploegman@middlebury.edu or 802-443-2382 and Jodi Litchfield can be reached at litchfie@middlebury.edu or 802-443-5936. All discussions will remain confidential.I will make best efforts to provide accommodations but, given the current circumstances, I can’t guarantee the requirements of the letter will be met.
You may need to find data for your blog post. I encourage you to go to the library and work with the librarians to find more resources, but be sure to cite what you use and provide your own insights. I highly recommend you reach out to Ryan Clement, the Economics reference librarian, as soon as you start settling on a topic. Ryan is a wealth of knowledge about economic data and literature. His email is rclement@middlebury.edu, and his website is at go.middlebury.edu/ryan.
I expect you to adhere to the Middlebury Honor Code. The Middlebury Honor Code is described at go/honorcode. You can collaborate on all assignments except the blog post (see the “Blog post” section under “More details on deliverables” for more details). If you collaborate with someone, you must acknowledge those you worked with and submit your own final writeup. For the blog post, an acknowledgements section at the back is a good way to do this.
It’s not a problem if you need a small extension (small = on the order of 24 hours), just send me an email or Slack DM letting me know that you’ll be submitting it a little later. If you need a longer extension, reach out. It’s generally going to be fine (I ask that you reach out so I know things are ok, or if they’re not, how I can help).
Sometimes life happens, and you aren’t able to submit things on time. That’s ok. As I mention elsewhere in this syllabus, I start from the premise that you’re here to learn. My preference is that you do all the work for this class (or as much as possible/feasible). Combining the two, it seems a bit odd and counterproductive to penalize you for late work—if you want more of a thing, why disincentivize it? The point is: if you weren’t able to do/submit an assignment for whatever reason, it’s still worth your time to do/submit it!
(In the past I’ve used late penalties due to fairness concerns around grading, but that’s not an issue under the labor-based contract for this class. “But won’t this just let people not do the work until the last minute with all the solutions available?” I mean yeah maybe but like… (a) cramming is an extremely poor learning strategy, (b) I trust you to put in your honest effort to learn, and (c) leaving everything till the last minute will probably end up being more work and stress than just doing things as they’re assigned.)
Since you won’t be penalized for effortful-but-incorrect answers, please don’t take extensions if you’re at the wall and unsure if your answers are correct. I don’t mind extensions, but this kind of extension usually won’t do much for your learning beyond maybe getting you frustrated. You’re almost surely better off just submitting your effortful attempts and reviewing the solutions carefully.
If you go by a different name or pronouns than what is listed on the roster, please let me know. I expect us to refer to each other by our preferred names and pronouns.
Sometimes people get sick (even without a pandemic in the background), or are otherwise unable to come to class/submit assignments for whatever reason. That’s ok. If you need to take time off for illness, a mental health break, or some other reason, please do so. You don’t need to let me know in advance, though advance notice is always appreciated. If you missed a longer stretch of classes/assignments (e.g. two weeks because of COVID) you can make up the assignments with no penalty. I’m always happy to discuss course content you’re struggling with, just send me an email/slack DM and we’ll go from there.
A brief note letting me know where you’re at and what I can do to support you is fine, even if it’s after the fact. I don’t need doctor’s letters or detailed explanations. My working assumption is that you want to be in class and learn and that if you’re not there it’s because something came up and you were either constrained or chose to do whatever was marginally most valuable/important to you. Either way, I respect your choice. If you choose not to share any details with me, that’s fine too. My goal to make sure you’re able to take time when you need it without seeking approvals or permissions and with the support you need to catch up. If you anticipate or experience a particularly long absence, please do reach out so I’m aware you’re ok (or not) and I can help you with any material you’re struggling with.
Please take care of yourself. College can be a demanding (and rewarding) experience. If you or one of your friends, peers, classmates is facing mental health problems, please consider seeking help. Resources include your Common’s Dean, Resident Life Staff, Middlebury College Counseling Center, and ULifeLine website.
The schedule below shows major assignments and due dates for the course. I will revise the schedule as/when things change. Unless specifically instructed otherwise, assume all assignments are due at midnight Eastern US time of the day they are listed as due.
The statistical concepts are listed in loose clusters by chapter/week and are meant to give you some guidance on what you can expect to learn. For example, the topics listed next to the first week will likely cover the first two weeks. Within each cluster they aren’t necessarily in order.
Key:
| Week | Week of | ISRS section | Statistical concepts | Lab topic | Monday | Tuesday | Wednesday | Thursday | Friday |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Sep 7 | 1.0-1.4 | Correlation vs causation, | Lab 0: Getting started (at home) | LR 2 assigned | PS 1 assigned | |||
| population vs sample, | Lab 1: Intro to R (lab time) | LR 1 due | |||||||
| sample statistics, | Black/White disparities in | ||||||||
| data and visualization. | mass incarceration | ||||||||
| 2 | Sep 14 | 1.5-1.7 | Lab 2: Intro to data | CQ 1 assigned | LR 2 due | PS 1 due | |||
| Black/White income gap | LR 3 assigned | PS 2 assigned | |||||||
| 3 | Sep 21 | 2.1-2.3 | Hypothesis testing, | Lab 3: Normal distribution | LR 3 due | PS 2 due | |||
| simulations, two-sided tests, | COVID-19 case rates and | LR 4 assigned | PS 3 assigned | ||||||
| inferential errors (1, 2, S, M), | income distributions | ||||||||
| 4 | Sep 28 | 2.4-2.8 | Normal distributions, | Lab 4: Intro to inference | CQ 1 due; | LR 4 due | PS 3 due | ||
| Central Limit Theorems, | and sampling distributions | CQ 2 assigned | LR 5 assigned | PS 4 assigned | |||||
| analytical vs simulation results. | Income inequality and | ||||||||
| public sentiment | |||||||||
| 5 | Oct 5 | 3.1-3.2 | 1 and 2-sample proportions, | Lab 5: Confidence intervals | LR 5 due | PS 4 due | |||
| Best predictors, | Climate change and | LR 6 assigned | PS 5 assigned | ||||||
| linear functions, two-way tables, | public sentiment | ||||||||
| joint/marginal/conditional distributions. | |||||||||
| 6 | Oct 12 | 4.1-4.3 | t-tests, bootstraps, | Lab 6: Inference for | CQ 2 due; | LR 6 due | PS 5 due | ||
| causal graphs, omitted variables, colliders. | categorical data | CQ 3 assigned | LR 7 assigned | PS 6 assigned | |||||
| 7 | Oct 19 | 4.4-4.5 | Lab 7: Inference for | LR 7 due | PS 6 due | ||||
| numerical data | LR 8 assigned | PS 7 assigned | |||||||
| 8 | Oct 26 | 5.1-5.3 | Linear models with error, | Lab 8: Linear regression | CQ 3 due; | LR 8 due | PS 7 due | ||
| regression coefficients and ATEs, | LR 9 assigned | PS 8 assigned | |||||||
| 9 | Nov 2 | 5.4 & 6.1 | regressions and hypothesis tests, | Lab 9: Multiple regression | LR 9 due | PS 8 due | |||
| regressions with dummy variables. | LR 10 assigned | PS 9 assigned | |||||||
| 10 | Nov 9 | Lab 10: Intro to Stata | CQ 4 assigned; | LR 10 due | PS 9 due | ||||
| LR 11 assigned | PS 10 assigned | ||||||||
| 11 | Nov 16 | Lab 11: t-tests and summary | LR 11 due | PS 10 due | |||||
| statistics in Stata | LR 12 assigned | PS 11 assigned | |||||||
| Break | Nov 23 | ||||||||
| 12 | Nov 30 | Lab 12: Regression in Stata | CQ 4 due | LR 12 due | PS 11 due | ||||
| Finals | Dec 7 | Final posting date | |||||||
| for blog posts |