Project outputs
This page lists the outputs of the project to date, and includes peer-reviewed publications, conference presentations and software. It will be updated regularly.
Peer reviewed journal articles
1. A Bayesian localized conditional autoregressive model for estimating the health effects of air pollution, by Duncan Lee, Alastair Rushworth and Sujit Sahu, Biometrics, 70 (2014), 419-429. A link to the paper is here.
2. Controlling for localised spatio-temporal autocorrelation in long-term air pollution and health studies, by Duncan Lee and Richard Mitchell, Statistical Methods in Medical Research, 23 (2014), 488-506
Conference presentations and seminars
1. A rigorous statistical framework for estimating the long-term health effects of air pollution, by Sujit Sahu, Duncan Lee and Richard Mitchell. A contributed talk given at the 2013 Annual UK review meeting on outdoor and indoor air pollution research, at Cranfield University. The conference proceedings can be found here.
2. A Localized Conditional Autoregressive Model for Residual Spatial Confounding in Air Pollution and Health Studies, by Duncan Lee, Sujit Sahu and Alastair Rushworth. An invited talk given at the 2013 Joint Statistical Meeting (JSM) in Montreal, Canada. The slides for the talk can be found here JSMslides.
3.Controlling for localised spatio-temporal confounding in long-term air pollution and health studies, by Duncan Lee and Richard Mitchell. An invited talk given at the 2013 GEOMED meeting in Sheffield. The slides for the talk can be found here GEOMEDslidesLee.
4. Investigating long-term health impacts of particulate matter in the UK: a novel spatio-temporal approach, by Alastair Rushworth, Duncan Lee, Richard Mitchell and Sujit Sahu. A contributed talk given at the 2013 GEOMED meeting in Sheffield. The slides for the talk can be found here GEOMEDslidesRushworth.
5. Statistical challenges in estimating the long-term health impact of air pollution, by Duncan Lee. An invited seminar at the Medical University of South Carolina. The slides for the talk can be found here LeeMUSC.
6. Statistical challenges in estimating the long-term health impact of air pollution, by Duncan Lee. An invited seminar at the University of Southampton. The slides are largely the same as for the talk above.
7. Extending conditional-autoregressive models for space-time disease mapping, by Alastair Rushworth and Duncan Lee. An invited talk given at the 2014 International Society of Bayesian Analysis meeting in Cancun. The slides for the talk can be found here ISBARushworth.
8. Air quality modelling for health impacts studies, by Paul Agnew. An invited talk given at the 2014 Royal Statistical Society conference in Sheffield. The slides for the talk can be found here RSSAgnew.
9. Hierarchical Bayesian spatio-temporal modelling of air pollution in the UK for estimating long-term exposure, by Sabyasachi Mukhopadhyay, Sujit Sahu and Lucy Neal. An invited talk given at the 2014 Royal Statistical Society conference in Sheffield. The slides for the talk can be found here RSSMukhopadhyay.
10. Gaussian Markov random field models for capturing localized residual spatio-temporal confounding in air pollution and health studies, by Alastair Rushworth and Duncan Lee. An invited talk given at the 2014 Royal Statistical Society conference in Sheffield. The slides for the talk can be found here RSSRushworth.
11. Statistical challenges in estimating the long-term health impact of air pollution, by Duncan Lee. An invited talk given at the Lancaster University. The slides for the talk can be found here Leelancaster.
12. Two-stage spatio-temporal modelling of the health effects of air pollution in England, by Alastair Rushworth. An invited talk given at the 2015 Spatial Statistics conference in Avignon. The slides for the talk can be found here Rushworthavignon.
13. Challenges in modelling air pollution and understanding its impact on human health, by Alastair Rushworth. An invited talk given at the 2015 Joint Statistical Meeting in Seattle. The slides for the talk can be found here RushworthJSM.
14. On generating a flexible class of anisotropic spatial models using Gaussian predictive processes, by Sujit Sahu. An invited talk given at the 2015 International Statistical Institute conference in Rio. The slides for the talk can be found here SahuISI.
15. Quantifying the long-term effects of air pollution on health in England using space time data, by Duncan Lee. An invited talk given at the 2015 GEOMED meeeting in Florence. The slides can be found here LeeGEOMED2015,
16. Using small-area spatial data for statistical and epidemiological reseaerch, by Duncan lee. An invited talk given to the Scottish Government in Edinburgh. The slides can be found here LeeScottishgov.
Software
1. CARBayesST - An add-on package to the statistical software R to implement a class of spatio-temporal generalised linear mixed models for areal unit data, with inference in a Bayesian setting using Markov chain Monte Carlo (McMC) simulation. The spatio-temporal autocorrelation is modelled by random effects, which are assigned conditional autoregressive (CAR) prior distributions. A number of different random effects structures are available, and full details are given in the vignette accompanying the package. The software was developed to model disease data as part of this project and is available on the Comprehensive R Archive Network (CRAN) here.