Course description
Spatial and Spatial-Temporal Data Analysis using Bayesian Hierarchical Models
This training course will introduce powerful Bayesian spatial and spatial-temporal modelling techniques that enable researchers to analyse small area spatial-temporal data arising in the social, economic, political and public health sciences.
The course will provide not only the underlying statistical theory for data analysis but also the hands-on experience necessary to apply different spatial and spatial-temporal techniques to visualise and analyse a wide variety of practical datasets.
Do you work at this organisation and want to update this page?
Is there out-of-date information about your organisation or courses published here? Fill out this form to get in touch with us.
Upcoming start dates
Suitability - Who should attend?
Entry Requirements
The course is aimed at PhD students, Post-Doctoral researchers and academic staff in the social, economic, political and public health sciences who have a background in quantitative methods at least up to the standard linear regression model. Prior experience of generalized linear models will be helpful for some parts of the course as will some familiarity with the basic ideas behind Bayesian inference.
Outcome / Qualification etc.
You will gain a broad knowledge of the diversity of current approaches to modelling small area spatial and spatial-temporal data. Through practical sessions, you will acquire hands-on experience in analysing spatial and spatial-temporal data arising in different fields. Upon completion of the course, you should be able to:
1. use relevant knowledge to analyse small area spatial and spatial-temporal datasets, from exploratory analysis to model development, and to produce relevant model outputs for answering substantive questions.
2. manage, manipulate and visualise spatial and spatial-temporal data using R and implement Bayesian spatial and spatial-temporal models via WinBUGS.
3. recognise issues and challenges presented in a spatial/spatial-temporal dataset with the aim to inform future model development, via modifying existing methodology and/or developing new modelling techniques.
Training Course Content
- Types and properties of spatial and spatial-temporal data and implications for model building;
- Techniques for testing spatial heterogeneity and autocorrelation; clusters and hotspot detection;
- Uses of maps and graphics for visualization using R;
- Bayesian inference for spatial and spatial-temporal data;
- Introduction to Markov chain Monte Carlo methods;
- Simple Bayesian regression models;
- Bayesian hierarchical models for spatial data, both continuous (e.g., normal/log normal data) and discrete (e.g., count/binary data);
- Implementation of various Bayesian hierarchical models in WinBUGS for spatial data;
- Introduction to spatial econometric models;
- Regression diagnostics with particular relevance for spatial data;
- Introduction to modelling small area time series data, including linear models in time, autoregressive models and interrupted time series models;
- Introduction to various strategies/structures for modelling spatial-temporal data;
- Discussion of space-time models using practical examples.