|Delivery Type||Delivery length / details|
|Lecture||22 x 1 Hour Lectures|
|Assessment Type||Assessment length / details||Proportion|
|Semester Exam||2 Hours (Written Examination)||100%|
|Supplementary Exam||2 Hours (Written Examination)||100%|
On completion of this module, a student should be able to:
1. construct and interpret Classical confidence intervals and tests of hypotheses for a population mean (Normal) and a probability parameter;
2. set up a Bayesian analysis of the same situations;
3. interpret prior and posterior distributions for parameters and construct Bayesian confidence intervals;
4. explain the differences between Classical and Bayesian analyses;
5. extend the ideas to other distributional families.
This module re-examines the ideas of likelihood, confidence intervals and hypothesis testing in Classical Inference and considers their interpretation more deeply. An alternative approach known as Bayesian Inference is introduced in which prior information is modelled in the form of a distribution and updated in the presence of data using Bayes's Theorem. The concepts prior, posterior, predictive and preposterior are introduced. Applications to inferences about a (Normal) population mean, a (Binomial) probability parameter and other distributional families are discussed in detail. The meanings and interpretations of the two approaches are discussed at length.
To introduce the basic ideas and concepts of statistical inference.
2. BAYESIAN INFERENCE Bayes' Theorem. Prior and posterior odds. Prior and posterior distributions. Conjugate families. Prior knowledge and prior ignorance. Quantification of knowledge. Predictive distributions. Preposterior distributions. Bayesian point estimation, loss functions.
3. CONFIDENCE STATEMENTS Classical: pivotal functions, confidence intervals. Bayesian: highest density intervals, predictive intervals. Interpretation of relative likelihood intervals.
4. HYPOTHESIS TESTING Classical: null and alternative hypotheses. Neyman Pearson theory. UMP tests.
5. OVERVIEW Comparisons between Classical and Bayesian approaches.
This module is at CQFW Level 6