Course Overview
Bayesian methods are widely used to analyse complex data structures, including population counts, time-use data, and other non-standard outcomes. This course introduces both basic and intermediate Bayesian regression methods, with a strong emphasis on practical implementation using R and Stan. The course is designed for participants with prior experience in regression modelling who wish to develop the skills and confidence to carry out Bayesian estimation, generate predictions, and communicate uncertainty through clear visual summaries, including for hierarchical and spatially structured parameters. Participants will study generalized linear models (including logistic regression) and multilevel/hierarchical regression frameworks. Teaching sessions are complemented by a participant-led project, culminating in poster presentations on the final day of the course. An open and reproducible research approach is adopted throughout, with shared code repositories used to support learning, collaboration, and dissemination. This course is brought to you by a team affiliated with the Cathie Marsh Institute for Social Research (CMI) at the University of Manchester.
Workshop Leads
Professor Wendy Olsen Wendy Olsen is Senior Statistician at the Office for National Statistics (UK). She was Professor of Socio-Economic Research at the University of Manchester. Her research spans labour markets, poverty, inequality, and advanced quantitative methods. She does applied regression modelling and uses Bayesian approaches. She has extensive experience delivering advanced methods training.
Contact: wendy.olsen@manchester.ac.uk
Dr Diego Andrés Pérez Ruiz Diego Pérez Ruiz is a Lecturer in Social Statistics at the University of Manchester. His expertise includes Bayesian statistics, survey methodology, regression modelling, and reproducible research. He teaches across undergraduate and postgraduate progrichmes and is actively involved in applied and methodological research using Bayesian and computational approaches.
Contact: diego.perezruiz@manchester.ac.uk
Course Objectives
By the end of this course, participants will be able to:
- Specify, estimate, and interpret regression models within a Bayesian framework using R and Stan.
- Apply generalized linear models (including logistic, Poisson, and ordinal models) to real-world data.
- Understand and diagnose uncertainty in parameter estimates using posterior distributions and Bayesian model comparison tools.
- Implement hierarchical and multilevel regression models, including spatial extensions where appropriate.
- Produce reproducible, well-documented analytical outputs suitable for research dissemination and professional reporting.
Preparatory work or pre-requisites
Prior to the beginning of the course, it would be useful to access and comprehend some of, or most of, the two books listed below.
- Jones, Elinor, Simon Harden and Michael J. Crawley (2022), The R Book, 3rd ed., London: Wiley.
- Dalgaard, Peter (2008) Introductory Statistics with R. NY: Springer, URL
Essential: Then read part of the technical article by Bürkner (2017) (pages 6-17, available open access), and study our preparatory pack which we will circulate the week before the course.
- Bürkner, Paul (2017), brms: An R Package for Bayesian Multilevel Models Using Stan, Journal of Statistical Software, 80:1 (note this article also has code in a repository which we will use during the course).
Who should attend?
This course is suitable for participants working or researching within data science, social statistics, and related quantitative fields, in either academic or applied settings. It is particularly appropriate for analysts, data scientists, and researchers who already use regression models and wish to extend their skills to Bayesian estimation, uncertainty quantification, and hierarchical modelling. Participants who primarily use software such as Stata or SAS are very welcome; the course demonstrates how Bayesian regression techniques implemented in R and Stan relate directly to familiar workflows.
Teaching Approach
- Each morning and afternoon, we combine a short lecture with question and answer coaching.
- We will study the provided worksheets and you will receive one summary ‘cheat sheet’ each day. This has a list of formulas and sample code.
- A daily practical hour includes starting assignments which keen participants could continue in the evening.
- PhD students are very welcome in our summer school. Those without a project can use the data we provide, for free.
Course Timetable
This course will take place in-person Monday 6 July - Friday 10 July
Monday: 1:30–5:00pm
- Foundations of Bayesian Regression & MCMC
- Review of key regression concepts (GLMs, model fit, likelihood)
- Introduction to Bayesian regression and Bayes’ theorem
- Markov Chain Monte Carlo (MCMC) methods
- Practical session using R (Stan and related packages)
- Project topic selection and mini-project launch
Tuesday: 9:00–12:30pm
- Bayesian Logistic Regression
- Logistic (logit) models for binary outcomes
- Interpretation of Bayesian coefficient distributions
- Model fit and comparison (AIC, BIC)
- Hierarchical logistic models
- Practical implementation in R / Stan
Tuesday: 1:30–5:00pm
- Project Work & Applied Practice
Wednesday: 9–12:30pm
- Bayesian Models for Counts and Non-Negative Outcomes
- Poisson and related count models
- Overdispersion and offsets
- Applications to population counts and expenditure data
- Model comparison and inference
Wednesday: 1:30–5pm
- Applied Modelling & Project Development
Thursday: 9–12:30pm
- Spatial and Multilevel Bayesian Models
- MCMC methods (Gibbs, Metropolis–Hastings)
- Spatial autocorrelation
- Multilevel and spatial models (e.g. BYM2)
- Prediction and visualisation
Thursday: 1:30–5pm
- Advanced Applications & Project Support
Friday: 9–12:30pm
- Ordinal and Multilevel Regression + Participant Presentations
- Ordinal regression models
- Multilevel modelling and sample size considerations
- Mean Squared Error (MSE) and simulation approaches
- Participant project presentations and discussion
Please note: A more detailed timetable, preparatory materials, and software guidance will be provided to participants ahead of the course and upon registration.
What's included in the course?
Each full day includes a vegan buffet lunch served 12.30-1.30pm. There are morning and afternoon refreshment breaks with tea, coffee, water, and pastries/cakes.
The course includes a social programme - these are optional but free social events for everyone attending our summer school to meet attendees from other courses in a relaxed environment.
Accommodation and travel are not included in the course price. You will need to arrange any accommodation and travel separately.
Cost
- Full price: £900
- PGR/Reduced Rate: £600
As well as PGRs, reduced fees are available to those working within the voluntary, charity and community sector. We also have two bursary options available for those entitled to reduced fees. Please view more information on our main Summer School website to find out more and how to apply.
Book Your Place
Please purchase via our online store below before 15 June (payment by card only). If you any questions, or will have trouble purchasing by this date please get in touch with methods@manchester.ac.uk.
Any questions, please do not hesitate to contact us on methods@manchester.ac.uk
Credits:
Created with images by David - "Group of confident business people point to graphs and charts to analyze market data, balance sheet, account, net profit to plan new sales strategies to increase production capacity." • whyframeshot - "Business meeting showcases agile brainstorming with data analysis charts on tablets.,Financial growth prediction strategy through predictive analytics and data-driven insights for strategic business"