Regression uses data to study mathematical relations among two or more variables, with the purpose of understanding trends, identifying significant predictors, and forecasting. The course covers simple and multiple regression, the method of least squares, analysis of variance, model building, regression diagnostics, and prediction. Students estimate and test significance of regression slopes, evaluate the goodness of fit, build optimal models, verify regression assumptions, suggest remedies, and apply regression methods to real datasets using statistical software.
This course satisfies the AU Core Integrative Capstone Requirement for the Statistics BS, having you build on your prior statistical knowledge to apply it in the new context of linear modeling. Also as part of the capstone experience, this course will require you to identify and execute a significant project that you will work on throughout the semester.
You will use Canvas (https://american.instructure.com/) to obtain and turn in projects and homework assignments. Students will use their AU credentials to log in to Canvas. AU’s Canvas Support team recommends using the latest version of Chrome or Firefox to optimize your experience. You can get help by using the Help menu located at the bottom of Global Navigation after you log in. Please also consult the global Canvas Community student guide for an explanation of key tools and features.
All lecture material will be posted to my GitHub Pages website: (https://dcgerard.github.io/stat_415_615/).
Required: Applied Linear Statistical Models (Fifth Edition) by Kutner, Nachtsheim, Neter, and Li. They also released a book called “Applied Linear Regression Models (Fourth Edition)” which is just a subset of this book, and this is fine too. I will generally abbreviate this book as KNNL.
Supplemental: The free Hands-on Programming with R provides a crash course in the basics of the R programming language.
Optional: The free Regression and Other Stories (ROS). Code and data for this book are available here as well. It’s one of the best applied books I’ve read on applied regression analysis. The only issues are that the code they use is a little non-standard, and it is not rigorous enough for a Statistics program. But check it out!
Optional: The Statistical Sleuth. This is the best non-mathematical treatment of basic Statistical analysis that exists. I use some of their datasets, and reference this book for best practices. If you do not remember basic statistical concepts, this is the best book you can get to review those concepts.
I will also provide you with supplemental PDF readings as required.
We will use the R computing language to complete assignments. R is free and may be downloaded from the R website (http://cran.r-project.org/). In addition, I highly recommend you interface with R through the free RStudio IDE (https://www.rstudio.com/). R and RStudio are also available on computers in the Anderson Computing Complex, the Center for Teaching, Research, and Learning Lab (CTRL) in Hurst Hall, in addition to various labs across campus. R Studio may also be run from your web browser using American University’s Virtual Applications System. Please see me during office hours if you have questions regarding R.
Usual grade cutoffs will be used:
Grade | Lower | Upper |
---|---|---|
A | 93 | 100 |
A- | 90 | 92 |
B+ | 88 | 89 |
B | 83 | 87 |
B- | 80 | 82 |
C+ | 78 | 79 |
C | 73 | 77 |
C- | 70 | 72 |
D | 60 | 69 |
F | 0 | 59 |
Individual assignments will not be curved. However, at the discretion of the instructor, the overall course grade at the end of the semester may be curved.
This is an AU capstone course, where the final project is a vital component to the AU core curriculum. As such, if you fail the final project, then you fail the course, even if your aggregate grade is above the fail level.
Your final project grade will also be adjusted based on anonymous peer assessment of your contribution to the project. If all your colleagues say that you didn’t do anything, then you will get no credit for the project.
If you are taking this class for graduate credit, I am going to require you to learn the linear algebra behind linear regression.
I am going to have a recorded lecture on linear algebra that you will be required to watch.
I will give you separate homework and exam questions that use linear algebra in linear regression (e.g. create design matrices, implement linear regression in R using matrix algebra, etc)
*02/27/2024: Section 001: Midterm 1
*02/28/2024: Section 002: Midterm 1
03/01/2024: Dune: Part Two release date.
03/12/2024: Section 001: Spring break. No class.
03/13/2024: Section 002: Spring break. No class.
*04/23/2024: Section 001: Midterm 2
*04/24/2024: Section 002: Midterm 2
04/30/2024: Section 001: Spring study day. No class.
05/01/2024: Section 002: Spring study day. No class.
05/07/2024: Section 001: Present final projects during final exam period (By Zoom).
05/08/2024: Section 002: Present final projects during final exam period (By Zoom).
*These dates are subject to change.
The learning objective of this course is to give you the main concepts and a working knowledge of regression techniques that are routinely used to analyze different types of data. At the end of this course, you are expected to be able to:
At the discretion of the faculty member and before the end of the semester, the grade of I (Incomplete) may be given to a student who, because of extenuating circumstances, is unable to complete the course during the semester. The grade of Incomplete may be given only if the student is receiving a passing grade for the coursework completed. Students on academic probation may not receive an Incomplete. The instructor must provide in writing to the student the conditions, which are described below, for satisfying the Incomplete and must enter those same conditions when posting the grades for the course. The student is responsible for verifying that the conditions were entered correctly.
Conditions for satisfying the Incomplete must include what work needs to be completed, by when the work must be completed, and what the course grade will be if the student fails to complete that work. At the latest, any outstanding coursework must be completed before the end of the following semester, absent an agreement to the contrary. Instructors will submit the grade of I and the aforementioned conditions to the Office of the University Registrar when submitting all other final grades for the course. If the student does not meet the conditions, the Office of the University Registrar will assign the default grade automatically.
The Associate Dean of the Academic Unit, with the concurrence of the instructor, may grant an extension beyond the agreed deadline, but only in extraordinary circumstances. Incomplete courses may not be retroactively dropped. An Incomplete may not stand as a permanent grade and must be resolved before a degree can be awarded.
Standards of academic conduct are set forth in the university’s Academic Integrity Code. By registering for this course, students have acknowledged their awareness of the Academic Integrity Code and they are obliged to become familiar with their rights and responsibilities as defined by the Code. Violations of the Academic Integrity Code will not be treated lightly and disciplinary action will be taken should violations occur. This includes cheating, fabrication, and plagiarism.
I expect you to work with others and me, and I expect you to use online resources as you work on your assignments/projects. However, your submissions must be composed of your own thoughts, coding, and words. You should be able to explain your work on assignments/projects and your rationale. Based on your explanation (or lack thereof), I may modify your grade.
If you use an online resource, please cite it with a URL. This includes any generative AI source (e.g. ChatGPT, Bard, or Copilot). Failure to include a URL citation for an online resource will be considered a violation of the Academic Integrity Code.
If you do not understand an online resource, but believe it to be useful for a project/assignment, please ask me for help.
It is a violation of the Academic Code of Integrity if you obtain past homework solutions from students who took the course previously (whether they wrote those solutions, or I wrote those solutions).
All solutions that I provide are under my copyright. These solutions are for personal use only and may not be distributed to anyone else. Giving these solutions to others, including other students or posting them on the internet, is a violation of my copyright and a violation of the student code of conduct.
This syllabus is a guide for the course and is subject to change with advanced notice. These changes may come via Canvas. Make sure to check Canvas announcements regularly. You are accountable for all such communications.