|
|
|
This is the preliminary program for the 2007 Joint Statistical
Meetings in Salt Lake City, Utah.
|
|
|
The views expressed here are those of the individual authors and not necessarily those of the ASA or its board, officers, or staff. Back to main JSM 2007 Program page |
= Applied Session,
= Theme Session,
= Presenter| CE_01C | Sat, 7/28/07, 8:30 AM - 5:00 PM | CC-151 G |
| Bayesian Modeling in Practice - Continuing Education - Course | ||
|
ASA |
||
| Instructor(s): David Spiegelhalter, MRC Biostatistics Unit | ||
| There has been an extraordinary growth in the application of Bayesian methods in diverse fields - from fisheries to the regulation of medical devices, from mapping diseases to court cases. There are many articles, textbooks and some reasonable software, so that if you know what you want to do you can generally do it. However anyone wanting to conduct a Bayesian analysis is faced with choices for which rigid rules cannot be specified: for example, what type of model might be reasonable in this context, what kind of prior assumptions are reasonable, how to implement the analysis in software, how to monitor and check the computations, how to communicate the results, how to assess the importance and plausibility of all the assumptions and, most important, how to convince a skeptical audience that the analysis is reasonable? This course will deal with the practical aspects of Bayesian modelling, not only in terms of actually getting results out, and also how to deal with these difficult strategic questions. A wide range of applications will be used to illustrate the implementation of models of varying complexity, and we shall try and provide some tentative guidance on some of the more tricky aspects of Bayesian analysis. A particular aim is to show how we can acknowledge and explicitly include in a model realistic assumptions about potential inadequacies in the available data. The examples will be illustrated using the WinBUGS software. Prerequisite: We shall assume familiarity with basic probability theory and common distributions, as well as standard statistical concepts such as linear and non-linear regression, generalized linear models, random-effects modelling including generalized linear mixed models, meta-analysis, residual analysis, model comparison using deviance, and so on. A basic familiarity with Bayesian methods would be useful but not essential, and it would help considerably if you had installed WinBUGS and at least run through the tutorial. The most important prerequisite is an enthusiasm for carrying out statistical analysis that adapts to the evidence available, rather than forcing data into whatever standard packages can provide. | ||
|
JSM 2007
For information, contact jsm@amstat.org
or phone (888) 231-3473. If you have questions about the Continuing Education program,
please contact the Education Department. |