PR stats still have some places on their Bayes GLMM course. Perfect for marine mammal researchers. Bayesian Multilevel Modelling Using *brms* for Ecologists (BMME01) www.prstats.org/course/bayesian-multilevel-modelling-using-brms-for-ecologists-bmme01 <https://www.prstats.org/course/bayesian-multilevel-modelling-using-brms-for-ecologists-bmme01?utm_source=chatgpt.com> *Dates:* October 20–24 and 27–31, 2025 *Format:* Live online course (4 hours per day) Also recorded with a further 30 days access *Fee:* £450 (standard) | £400 (early bird for the first 5 registrations) SOLD OUT Why This Course Is Ideal for Marine Mammal Researchers Marine mammal research often involves *complex, hierarchical, and data-limited systems*, making it a perfect field for Bayesian multilevel modelling. Whether studying population dynamics, spatial distribution, behavioural ecology, or habitat use, marine mammal data commonly include multiple sources of variability—across individuals, sites, seasons, and time periods. This course provides the statistical framework and practical tools to address these challenges directly. *1. Handling Hierarchical and Nested Data* Marine mammal studies often involve repeated observations of individuals within groups, colonies, or geographic regions. Traditional statistical approaches can struggle with this kind of nested data, leading to pseudoreplication or biased inference. Multilevel (hierarchical) Bayesian models allow you to explicitly model variation at each level—individual, group, site, or species—while appropriately sharing information across them. *2. Dealing with Sparse or Unbalanced Data* Marine fieldwork can be logistically challenging, leading to unbalanced sampling designs, small sample sizes, or missing data. Bayesian inference provides a natural way to incorporate uncertainty and make robust estimates even when data are limited or unevenly distributed. *3. Modelling Detection and Observation Processes* Acoustic monitoring, photo-identification, and aerial surveys introduce observation errors and detection probabilities that can bias results if not properly modelled. Bayesian methods make it straightforward to include detection submodels and propagate uncertainty through the full inference process. *4. Incorporating Spatial and Temporal Structure* Marine mammal ecology is inherently spatial and temporal: migration routes, foraging ranges, and environmental drivers vary in space and time. The *brms* framework allows users to include spatial and temporal random effects, model autocorrelation, and integrate environmental covariates such as sea surface temperature or prey density. *5. Integrating Diverse Data Sources* Bayesian models excel at combining information from multiple datasets—such as telemetry, acoustic detections, and visual surveys—into a single, coherent analysis. This flexibility makes it easier to synthesise data from different monitoring programs or species. *6. Communicating Uncertainty Transparently* Conservation and management decisions for marine mammals often rely on uncertain data. Bayesian credible intervals and posterior predictive checks help researchers communicate this uncertainty clearly and defensibly to policymakers and stakeholders. *7. Direct Application to Marine Mammal Research Questions* By the end of the course, participants will be able to construct and interpret models directly relevant to marine mammal science, such as: - Estimating population abundance or occupancy across survey areas - Assessing habitat preferences and environmental predictors - Modelling survival, movement, and behavioural states - Quantifying effects of noise, disturbance, or climate variables ------------------------------ COURSE DESCRIPTION Bayesian methods are rapidly becoming the standard for ecological data analysis. They offer a flexible, transparent, and robust approach to dealing with uncertainty, hierarchical data structures, and complex ecological processes. If you work with ecological data and want to extend your modelling skills beyond classical statistics, this course will give you the tools and understanding to do so with confidence. *Bayesian Multilevel Modelling using brms for Ecologists (BMME01)* is an intensive 10-day live online course designed to provide ecologists, environmental scientists, and applied researchers with a solid grounding in Bayesian hierarchical modelling using the *brms* package in R. Over the course of ten sessions, participants will be guided step by step through the key concepts and practical applications of Bayesian inference, with a strong focus on real ecological examples. The course combines clear, structured teaching with extensive hands-on coding and interpretation. Course Overview Participants will begin by reviewing the fundamentals of Bayesian inference—priors, posteriors, and credible intervals—before moving on to the construction and interpretation of generalised linear and multilevel models. You will learn how to fit and diagnose models for a range of ecological data types, including continuous, count, binary, and zero-inflated data, as well as how to address common challenges such as spatial and temporal autocorrelation. Throughout, you will work with real-world datasets and gain practical experience in coding models using *brms*, an R package that provides a user-friendly interface to *Stan*, one of the most powerful Bayesian computation frameworks available. By the end of the course, you will have the skills to build, diagnose, and interpret complex Bayesian models relevant to your own ecological research. Key Topics Include - The foundations of Bayesian inference and model formulation - Understanding priors, posteriors, and credible intervals - Generalised linear models (GLMs) and extensions to hierarchical models - Random effects, nested structures, and partial pooling - Modelling count, binary, zero-inflated, and multivariate ecological data - Incorporating spatial and temporal structures - Model checking, convergence diagnostics, and posterior predictive checks - Interpreting and communicating uncertainty in ecological research Who Should Attend This course is aimed at ecologists, environmental scientists, conservation biologists, statisticians, and postgraduate researchers who already have a working knowledge of R (data import, manipulation, and basic plotting) and want to develop practical expertise in Bayesian multilevel modelling. No prior experience with Bayesian methods is required—concepts and techniques are introduced progressively, with plenty of guided practice and individual feedback. Why Choose This Course - Expert instruction from *Dr Niamh Mimnagh*, an experienced statistical ecologist and educator - A carefully balanced mix of lectures, demonstrations, and hands-on coding - Small-group format encouraging discussion and individual support - All course materials, R scripts, and datasets provided - Access to recorded sessions for 30 days after the course - Certificate of attendance upon completion Registration Places are limited to ensure an interactive learning experience. The early bird rate of £400 applies to the first five registrants. Standard registration is £450. For full details, schedule, and registration, visit: www.prstats.org/course/bayesian-multilevel-modelling-using-brms-for-ecologists-bmme01 <https://www.prstats.org/course/bayesian-multilevel-modelling-using-brms-for-ecologists-bmme01?utm_source=chatgpt.com> -- Oliver Hooker PhD. PR stats