Bayesian Bootcamp: Fundamentals, Extensions, and Guidelines for Best Practice
Dr. Sarah Depaoli (www.sarahdepaoli.com)
Dr. Haiyan Liu (https://sites.google.com/view/ucmhaiyanliu)
Short Course Description:
This course does not assume any previous experience with Bayesian statistical modeling. It is intended to provide participants with a theoretical introduction, as well as provide practical examples and demonstrations. The Bayesian methodology will be described in relation to traditional frequentist approaches in order to provide participants with an understanding of the similarities and differences between the methods. The course starts with describing the philosophical and probability underpinnings of Bayesian and frequentist statistics, which acts as a springboard into Bayesian statistical modeling.
Participants should have experience with frequentist approaches to statistics (e.g., hypothesis testing, confidence intervals, least-squares, and likelihood estimation). Experience with statistical models through multiple regression is also desired. There are no requirements, but the participant’s experience in this workshop will be enhanced by previous statistical coursework or experience with advanced modeling concepts (e.g., general linear modeling, multivariate models), as well as familiarity with basic probability theory (e.g., joint, marginal, and conditional distributions, independence).
The principles of Bayesian statistics will be introduced using familiar concepts and simple examples (e.g., binomial models). Then we will move to simple linear regression and multiple regression. We will cover many aspects of Bayesian statistical modeling, including model construction, graphical representations of models, Markov chain Monte Carlo (MCMC) estimation and related concepts, evaluating hypotheses and model fit, and model comparison. We will also highlight specific “dangers” that users of Bayesian methods should be aware of.
Bayesian modeling has proved advantageous in many disciplines, but the examples used in this workshop will be primarily drawn from the social sciences. Input, output, and data will be provided for all examples covered in the workshop. Throughout the course, participants will be able to practice exercises using R and JAGS. Participants are encouraged to have a set-up where they can work on their own computer and also access the workshop video stream (e.g., two laptops set side-by-side, or using a multiple desktop feature of a single computer). Participants will be instructed on how to download free versions of the software prior to the course, and they will be provided with any relevant coding and data files as necessary.
Graduate students, emerging researchers, continuing researchers
Participants do not need any prior knowledge of Bayesian statistical modeling. Participants should have a basic understanding of conventional (frequentist) approaches up through multiple regression. Previous experience with more advanced models is a plus, but it is not required.
All models and exercises will be done using R and JAGS software. Participants will be provided all resources to perform these exercises on their own, and we will provide instructions on how to download free versions of the software prior to the course.
Dates and Times:
[Times provided in Pacific Standard Time]
April 30, 2021
9:00am – 12:00pm: Morning Session
12:00pm – 1:00pm: Lunch Break
1:00pm – 4:00pm: Afternoon Session
Participants will receive a code to use on their own computer to access a live-stream of the workshop, which will show the instructor, handouts, and software demonstrations. Electronic workshop materials will be provided ahead of time.
Participants may choose to watch live (synchronously), a recording (asynchronously), or both. Access to the workshops will be provided for one week after the end of the workshop. Note that asynchronous participation does not include real-time chat during the presentation, but a visual record of the chat will be provided in the recordings.
Content support is limited to real-time chat with synchronous participation. The instructor(s) and the online community (i.e., other participants) will participate in the chat to the best of their ability. However, it may not always be feasible for the instructor to monitor the chat feature because they will be busy with course content.
There may be times where the instructor presents hands-on activities, where the participants can follow along on their own computers. Instructions for software downloads (etc.) will be provided ahead of time. Support for such activities will be limited, given the online nature of the course.
How to Register:
Contact for Questions:
Sarah Depaoli (email@example.com
For course-related information:
Sarah Depaoli (firstname.lastname@example.org)
Haiyan Liu (email@example.com)
Dr. Sarah Depaoli is an Associate Professor of Quantitative Methods, Measurement, and Statistics in the Psychological Sciences Department at University of California, Merced. She teaches coursework covering advanced statistics, Bayesian modeling, longitudinal data analysis, and structural equation modeling. She is the author of the upcoming book, Bayesian Structural Equation Modeling, and her research has appeared in journals such as Structural Equation Modeling: A Multidisciplinary Journal, Multivariate Behavioral Research, Psychological Methods, Journal of Educational and Behavioral Statistics, and Nature Reviews. She has served on the editorial boards of several journals, has served as a guest editor for two journals, and currently serves as an Associate Editor for two journals: Multivariate Behavioral Research and Psychological Methods. She was the recipient of the 2011 Distinguished Dissertation Award from the American Psychological Association, Division 5, the 2015 Rising Star (Early Career) Award from the Association for Psychological Sciences (Quantitative Psychology division), and the 2020 Early Career Alumni Award from the University of Wisconsin, Madison. She is also an elected member of the Society of Multivariate Experimental Psychology (as of 2016, with 65 active members world-wide). Dr. Depaoli holds a B.A. in Psychology (Cal State Sacramento), an M.A. in Quantitative Psychology (Cal State Sacramento), and a Ph.D. in Quantitative Methods with a minor in Mathematical Statistics (University of Wisconsin, Madison).
Dr. Haiyan Liu is an Assistant Profess of Quantitative Methods, Measurement, and Statistics in the Psychological Science Department at the University of California, Merced. She teaches Research Method to undergraduate students and Multivariate Statistics and Introduction to Bayesian Data Analysis to graduate students. Her research interests lie in a broad range, including Social Network Modeling, Covariance Structure Analysis, and Bayesian Inference. She was the recipient of the 2019 Tanaka Award for Best Article in Multivariate Behavioral Research. Dr. Liu holds Ph.D. degrees in both Applied probability and Quantitative Methods (University of Notre Dame).