Statistics 451

Lecturer: Mr E Mullins

Date: 1995-96

Groups: JS/SS Mathematics, Two-subject moderatorship, JS/SS MSISS (elective)

Duration: 21 weeks

This course is concerned with the use of linear statistical models for the analysis of data from a variety of sources. The orientation is applied rather than theoretical, but such theory as is necessary for a proper understanding of the methods will be covered. Emphasis will be laid on data analysis as a model-building exercise and on the need to validate the assumptions of the models being used.

The course covers four broad areas:

1. Design of Experiments
2. Analysis of Variance
3. Multiple Regression Analysis
4. Logistic Regression Analysis

Data from a variety of disciplines will be analysed during the course. These will include: industrial experimentation, business forecasting, agricultural field trials, bioassay, epidemiology, environmental sciences, clinical trials. Students will be expected to participate actively in the course by analysing data between classes: a number of statistical packages will be used, including Minitab and Datadesk and possibly GLIM (depending on the availability of logistic regression facilities in Datadesk by early 1996).

References:

1. Box, G.E.P., Hunter, W.G. and Hunter, J.S., ``Statistics for Experimenters", Wiley, 1978.
2. Davies, O.L. (ed.), ``The Design and Analysis of Industrial Experiments", Longmans, 1978.
3. Neter, J., Wasserman, W. and Kutner, M.H. ``Applied Linear Statistical Models", Irwin, 1990.
4. Cox, D.R., ``Planning of Experiments", Wiley, 1992.
5. Box, G.E.P. and Draper, N.R., ``Empirical Model Building and Response Surfaces", Wiley, 1987.
6. Collett, D., ``Modelling Binary Data" Chapman and Hall, 1991.
7. Cox, D.R., and Snell, E.J., ``Analysis of Binary Data", Chapman and Hall, 1990.
8. McCullagh, P. and Nelder, J.A., ``Generalized Linear Models", Chapman and Hall, 1989.
9. Crowder, M.J. and Hand, D.J., ``Analysis of Repeated Measures", Chapman and Hall, 1990.