Level: BSc, MSci, MSc
Title: The analysis of designed experiments with multivariate data
Supervisor:

Dr Steve Coad
Research Area: Probability and Applications [Including Statistics]
Description:

This project involves a detailed study of multivariate analysis of variance (MANOVA), the main statistical technique used to analyse data from a designed experiment in which data consist of observations on more then one variable. In its simplest form, MANOVA can be used to compare several mean vectors, given random samples from multivariate normal distributions each with the same covariance matrix. This is known as one-way MANOVA. The technique extends naturally to more complicated experimental designs, such as randomised block and factorial designs. Other possible topics include repeated measures designs, profile analysis, simultaneous confidence intervals and methods for assessing multivariate normality.
Further Reading:
  • C. Chatfield and A. Collins, Introduction to multivariate analysis. Chapman & Hall, London, 1980.
  • R. Johnson and D. Wichern, Applied multivariate statistical analysis. Fourth edition. Pearson Prentice Hall, New Jersey, 1997.
Key Modules:

MTH5121 Probability Models

MTH6134 Statistical Modelling II
Other Information:
Current Availability: Yes