This course broadly covers the entire process of collecting, cleaning, visualizing, modeling, and presenting datasets. It has a MATH designation but is not a mathematics course. The focus is on applied statistics and data analysis rather than a detailed study of symbolic mathematics.
By the end of the semester you will feel confident collecting, analyzing, and writing about datasets from a variety of fields. You will be able to use these skills to address data-driven problems in a wide range of application domains.
- “The best thing about being a statistician is that you get to play in everyone’s backyard.”
I am an applied statistician. A sampling of projects and datasets I have worked on include:
- cell phone telemetry
- emergency room patient flow
- finding holes in specialized medical coverage in rural US
- Canadian court case citations
- Olympic figure skating scoring
- auto insurance risk factors
- 170k documentary photographs from the 1930’s
- treatment outcomes for open-angle glaucoma
- detecting radicalization from social media data
- financial warfare
We will talk about many of these projects throughout the semester.
Hans Rosling’s 200 Countries, 200 Years, 4 Minutes:
I have shown this to nearly all of my statistics courses, and while a bit dated it is still the best representation of what this course is all about.
Of course, most of our arguments will be in a written form. For an example of that, see this NYTimes article Gun homicides in New Zealand are about as common as deaths from falling from a ladder in the United States:
Notice how statistics is used as evidence in a larger arugment.
Or, if you like sports, here is the 538 article Steroids Probably Aren’t Causing Baseball’s Power Surge:
I’ll often give those of you interested in a particular subject (i.e., sports, economics, medicine) the chance to work with data that matches your interests.
Statistics in 209
This course is truly about the holistic process of doing statistics. There will be very little mathematical content. For that, you’ll need the MATH329 (Probability) and MATH330 (Statistics) sequence.
So, come into this course with an open mind about what we will be covering. I believe you will find it interesting, fun, and incredibly useful.
If we are not doing much mathematics, what will we be doing? We will be covering the basic elements of data analysis:
In order to actually do these things, we will work with the R programming language:
It is freely available for all major operating systems and is pre-installed on many campus computers.
To get an idea of the kinds of analysis we’ll be working on here are two projects from other applied statitistics courses I’ve taught:
I’ve avoided projects from other 209 courses because we may re-use some of those datasets. However, the basic structure of the projects, particularly the first one, is very similar.
In-Class Assessments (quizzes)
On most Tuesdays, there will be a short assessment covering the material from the prior week. Note that this includes both a conceptual understanding of the topics covered as well as the ability to apply these concepts to data with code. I will provide details on the class website with the exact topics on each assessment.
While the assessments serve to make sure you are following along with the general concepts, the core aim of the course is to teach you how to apply statistics to real-world questions. To this end, we will complete several data-oriented projects. These are the real heart of the course.
These projects consist of short written documents that mix code, graphics, and prose to provide a comprehensive analysis of a data set.
The in-class assessments are graded on a strictly pass/fail basis. The whole of these are converted into a letter grade according to how many you have passed (see syllabus for the conversion).
Your projects will receive a letter grade according to a rubric that I will distribute ahead of the due-date. The final grade will be determined by weighting the assessments and projects as follows:
- Projects: 67%
- Assessments: 33%
To pass the course, you must also miss no more than four class meetings. Attendance requires that you arrive on-time, complete any out of class assignments for the day, and fully engage with the course material. Failing to fulfill these attendance requirements may result in a failing grade for the course.
- Academic honesty: Cheating and plagiarism are grave scholarly offenses and potential grounds for expulsion; they are also a major barrier to your intellectual development. You are expected to familiarize yourself with the entirety of the University of Richmond’s Honor Code. If you are confused or unsure about appropriate citation protocol or any other aspect of the Honor code, please consult me before turning in an assignment.
- Special approval: If you have special approval forms for extra time on exams or any other circumstances I should know about, please speak with me as early as possible so that we can best accommodate your needs.
- Late work: You are expected to submit all work on-time. Late reports will be accepted after the due date with a full letter grade deduction for each 24 hour period it is late (rounded up).
- Attendance: You are expected to both attend and participate in most class meetings. If you must be absent due to illness or other pressing need, please let me know by email as soon as possible. A habit of arriving late, failing to participate, or failing to accomplish any out of class assignments is considered equivalent to an absence.
- Make-up work: In instances where students have a valid excuse for missing a quiz, please arrange to meet with me as soon as possible.
- Class conduct: During class I expect you to refrain from checking email, being on phones, or working on assignments for other classes.
- Computers: During programming assignments started in class, I expect you to use the computers in the lab. This is helpful for several reasons: it reduces distractions from iMessages and other materials on your laptop; all of the lab computers are configured using the same software and language set-up, reducing errors specific to your machine; and, other students and myself can share the same screen without worrying about modifying something on your personal machine.
- Office hours: If you would like to meet during my office hours, please just come by. No need to schedule an appointment. If you find me in my office at other times of the week, I am usually glad to meet then as well. Finally, I am also happy to make appointments outside of my normal office hours. These appointments are meant for discussing longer issues that are not appropriate for regular office hours (i.e., asking for recommendation letters or discussing an extended absence) or for students who cannot make my normal office hours. Please note that appointments should be booked at least 24 hours ahead of time.
- Email: I will also answer questions by email (it can, in fact, be much faster than scheduling an appointment for small issues). During the week, I aim to respond within 24 hours, with emails sent over the weekend responded to by Monday morning. If your question involves code, please attach your current lab or report as that will expedite my answering your question(s).
These address some of the most common questions and concerns that students have. If anything is unclear, please feel free to contact me for clarification at any point in the semester.
For the next class, please read the following short comment from the article “What is Statistics” by Emery N. Brown and Robert E. Kass. The original article is great but its written for academic statisticians who already know the field very well. The comments by David Madigan and Andrew Gelman (Columbia) are addressed to a more generalist audience:
We will discuss this on Thursday.