STA6349: Applied Bayesian Analysis - Fall 2024

Instructor Information

Dr. Samantha Seals
Email: sseals@uwf.edu
Office: 4/344

Meeting Times and Location

Day/time: MW 4:00-5:15 pm (Central)
Location: 52/163
Location: Zoom (see Canvas for link)

Office Hours

Monday: 1:00 - 3:00 pm (Central)
Tuesday: 1:00 - 3:00 pm (Central)
Thursday: 10:00 am - 12:00 pm (Central)
Other times by appointment

Class Websites

GitHub repository: https://github.com/samanthaseals/STA6349Fa24
GitHub website: https://samanthaseals.github.io/STA6349Fa24
Discord server: Applied Bayesian Analysis - Fall 2024 (see Canvas for link)

Course Description

This course is an applied first course in Bayesian data analysis. After reviewing basic probability theory, including marginal and conditional probabilities, Bayes’ Theorem will be reviewed and students will learn how to analyze data under the Bayesian framework using prior and data distributions to construct posterior distributions. While many probability distributions will be discussed, emphasis will be placed on the beta-binomial and normal-normal models. Basic simulations will be conducted to estimate posterior distributions and predict the probabilities of outcomes.

Topics

reproducible research using R/RStudio, Quarto, and GitHub
probability theory and Bayes theorem
probability distributions
prior, data, and posterior distributions
Beta-binomial model
conjugate families
normal-normal model
Markov chain Monte Carlo simulation of the posterior
making predictions using the posterior distribution

Student Learning Outcomes

demonstrate understanding of the differences between frequentist and Bayesian frameworks
distinguish between and use joint, conditional, and marginal probabilities
develop simple Bayesian models to answer research questions
use prior and data distributions to construct the posterior distribution
simulate values from posterior distributions and draw conclusions about the data
interpret data and simulation results in the context of the research question
disseminate analysis results to non-technical audiences

Course Materials

The following are books that we will be using directly in class:

Mathematical Statistics with Applications by Wackerly, Mendenhall, and Scheaffer.

Bayes Rules! An Introduction to Applied Bayesian Modeling by Johnson, Ott, and Dogucu.

The following texts are recommended as references, but not required. The links below take you to the electronic version of the text that is kindly provided to the public for free.

Probability by Kuter. (Note: this has been imported into Canvas thanks to the generosity of the author!)
Probability and Bayesian Modeling by Albert and Hu.

Course Format and Organization

This course will be taught synchronously via Zoom. You may join the classroom here: Zoom Classroom (see Canvas).

Note that all lectures will be recorded. If you choose to use the microphone to ask questions, your voice will be captured on the recording. If you do not feel comfortable with this, please do not use the microphone to ask questions. The chat will be monitored for questions and comments during lecture. Lecture videos will be posted to Canvas within two business days.

Breakout rooms will be utilized throughout lecture periods. Time in the breakout rooms will not be recorded.

Grading and Evaluation

The course grade will be determined as follows:

Activities (20%): Every few lectures, we will use the class time to work on an activity reflecting homework question(s) from the learning module. The resulting .html file will be submitted to the appropriate dropbox on Canvas. To allow for flexibility with life/work scheduling, activities will be due the following Sunday at 11:59 pm (Central). Note that the dropbox will lock at 8:00 am (Central) on Monday morning.
Projects - 2 (25% each): All projects will be completed using R and a report will be constructed using Quarto. Projects will be submitted as .html files to the appropriate dropbox on Canvas. To allow for flexibility with life/work scheduling, projects will be due the following Monday at 11:59 pm (Central). Note that the dropbox does not lock until the end of the semester.
Final Exam (30%): The final exam will be a written, closed book, proctored conceptual exam. While there may be some calculations needed, you will not be processing raw data on the proctored portion of the final exam. A scientific (not graphing) calculator will be allowed. The proctored final exam will be on Wednesday, December 4, 2024. Exams may not be taken late. Students will have 2 hours and 30 minutes to complete the exam.
- For those local to the Pensacola area, you may opt to come to the classroom for your final exam. When MathStat Proctoring reaches out, please let them know that you will be testing with me on campus. Our exam time is 2:00-4:30 pm in our scheduled classroom.
- For those requiring external proctoring, you must follow guidelines from MathStat Proctoring. You will receive an email from Proctoring the first week of the semester. Due to the number of students in our graduate programs, you must complete the form by September 13th. Otherwise, you will receive a 0 on the exam.

It is expected that all work submitted is the student’s own work. Helping each other with coding is encouraged, however, all written responses should be your own work. Evidence of academic dishonesty, including use of “homework help” websites (e.g., Chegg), use of AI (e.g., ChatGPT), and collaboration with other students, will be submitted to the Dean of Students. A grade of 0 will automatically be assigned for the assignment and there will be no opportunities to change that grade. If there are repeated incidents, the sanctions attached will increase in severity, including an F in the course and suspension from the university.

Late Policy

Projects 1 and 2 have due dates, however, the dropboxes will not close until the end of the semester. All students are automatically granted “extensions” for these projects.

Extensions are not available for the activities or the final exam.

Project Revisions

Projects 1 and 2 are eligible to receive two non-zero grades.

Note 1: dropboxes do not close until the end of the semester. You can resubmit at any time.
Note 2: if projects have formatting issues, I will leave a comment for you and issue a 0 grade solely for the purpose of Canvas notifying me when you’ve resubmitted. This 0 does not count towards the count of non-zero grades.

Course Grade

Final course grades will be determined according to the following scale. Conventional rounding rules will be applied.

Letter Grade	Weighted Score
A	93%–100%
A-	90%–92%
B+	87%–89%
B	83%–86%
B-	80%–82%
C+	77%–79%
C	73%–76%
C-	70%–72%
D+	67%–69%
D	60%–66%
F	< 60%

Important University Dates

Date	Event
Aug 19 (Mon)	Fall begins.
Aug 25 (Sun)	Drop/Add period ends.
Sep 2 (Mon)	Labor Day holiday - campus closed.
Nov 11 (Mon)	Veteran’s Day holiday - campus closed.
Nov 12 (Tues)	Withdrawal deadline (automatic grade of “W”).
Nov 28-29 (Thurs-Fri)	Thanksgiving holiday - campus closed.
Dec 6 (Fri)	Late withdrawal deadline (“W” or “WF”, see requirements below).

Students who are requesting a late withdrawal from class must have the approval of the advisor, instructor, and department chairperson (in that order) and finally, by the Academic Appeals committee. Requests for late withdrawals may be approved only for the following reasons (which must be documented):

A death in the immediate family.
Serious illness of the student or an immediate family member.
A situation deemed similar to categories 1 and 2 by all in the approval process.
Withdrawal due to Military Service (Florida Statute 1004.07)
National Guard Troops Ordered into Active Service (Florida Statute 250.482)

Requests without documentation should not be accepted. A request for a late withdrawal simply for not succeeding in a course does not meet the criteria for approval and should not be approved.

Additional Information for Students

Please see the University’s Confluence page for additional syllabus statements: https://confluence.uwf.edu/display/public/Additional+Syllabus+Statements