AS3C-Analysis of Structured Data

Module Provider: Applied Statistics
Number of credits: 20 [10ECTS credits]
Terms in which taught: Autumn and Spring
Module Convenor: Dr KL Ayres
Pre-requisites: AS2A AS2B
Co-requisites:
Modules excluded:
Module version for: 2008/9

Email: k.l.ayres@reading.ac.uk

Aims:
In many experiments or surveys, several different variables are recorded for each of many individuals. The problems associated with this sort of data can be tackled using multivariate analysis techniques. The methods to be discussed will be descriptive in nature and include the following topics: principal component analysis; canonical variates analysis; cluster analysis.
Many statistical techniques are only applicable when observations are independent. When successive observations on quantities such as weight or a measure of lung function are made the repeated measurements will usually be correlated. Statistical methods used in the analysis of this form of data will be described, such as the summary statistics approach, split-plot and multivariate analysis of variance. The use of mixed models for analysing repeated measurements will also be discussed.

Assessable learning outcomes:
By the end of the module it is expected that the student will have:

an appreciation of the role of multivariate analysis in statistics;

the ability to carry out commonly used techniques, such as principal components analysis and canonical variates analysis;

an awareness of repeated measurements and methods for analysing data in this form;

the ability to compare and contrast different approaches for analysing repeated measurements and be able to perform common types of analysis.

Additional outcomes:

Outline content:
MULTIVARIATE ANALYSIS
Graphical techniques to show multivariate data.
Principal Component Analysis.
Cluster Analysis.
Canonical Variate Analysis.
Correspondence Analysis.
Singular Value Decomposition.
Multi-dimensional scaling.
REPEATED MEASUREMENTS
Summary statistics.
Split-plot analysis of variance.
Multivariate analysis of variance.
Mixed models.
Use of SAS PROC GLM and PROC MIXED.
Recommended Reading:
Manly, B F J (1990). Multivariate Statistical Methods: A Primer. Chapman & Hall.
Krzanowski, W J (1988). Principles of Multivariate Analysis: A User's Perspective. Oxford Science Publications.

Brief description of teaching and learning methods:
Lectures supported by practicals.

Contact hours:

	Autumn	Spring	Summer
Lectures	9	18
Tutorials/seminars
Practicals	7	2
Other contact (eg study visits)	4
Total hours	20	20
Number of essays or assignments	1	1
Other (eg major seminar paper)

Assessment:
Coursework
Two assignments Weight: 30%
Penalties for late submission
Penalties for late submission of course work will be in accordance with University policy.
Examination
One paper of three hours duration Weight: 70%
Requirement for a Pass
An overall mark of at least 40%
Re-assessment
One examination paper of 3 hours duration