Exploratory Data Analysis in Archaeology

This module provides an introduction to the main exploratory multivariate techniques in use in archaeology.

This module provides an introduction to the main exploratory multivariate techniques in use in archaeology, namely cluster analysis, Principal Components Analysis (PCA), Correspondence Analysis (CA) and Discriminant Analysis. Extensions of these techniques, such as bootstrapped or detrended CA will also be explored. The module concentrates on the use of these techniques: what types of data they can examine, how to interpret the results, the pit-falls in the use of these techniques and so on. Wherever possible, the details of the underlying mathematics will not be examined in any detail.

Aims of the module

• A knowledge of the main multivariate statistical techniques used in archaeology.
• Practical experience of undertaking these analyses using the R statistical system.
• The relative strengths and weaknesses of the various techniques

Objectives of the module

• Choose and apply the appropriate techniques for various data sets and questions you may encounter.
• Be able to interpret the results of your analyses and identify potential problems.
• Report on your analyses in a appropriate manner.

Teaching Methods

Teaching will be by a mixture of lectures and supervised practical sessions. Classes will consist of two hours per week. Practical classes will normally involve direct supervision for one hour, often followed by a further hour during which time the tutor will be available to help as you work through exercises on your own.

You will also be given data sets to examine during the week between classes which will then be discussed at the start of the following week allowing you to gain practical experience in analysis and interpretation.

This module is assessed by means of a total of 5000 words of coursework consisting of a single data analysis project. You will be expected to identify a suitable data set with associated problems.

Module information

• Code: ARCL0087
• Credits: 15
• Coordinator: Kris Lockyear 
• Prerequisite: You should have a basic understanding of statistical methods, e.g., the normal distribution, probability, correlation. You should also be willing to master a command-line based statistical package.
• Handbook: open»

For registered students

• Reading list: open»

Availability

• Running in 2023-24