Day 1: Introduction
Day 2: Principal Component Analysis (PCA)
Day 3: Correspondence Analysis (CA)

Course Instructor

Georg Roth (Institute of Prehistoric Archaeology, Freie Universität Berlin)

Language

English

Objective

In the humanities, digital data collection increasingly results in large data tables where archaeological units are described by many attributes simultaneously – for instance, the counts of various types per assemblage. This allows for the arrangement of such units based on their overall similarity across all attributes, providing previously unattainable insights into complex relationships between the units. The course teaches the practical application of two standard methods (PCA and CA) for multivariate similarity-based ordination(s). Upon completion, participants will be able to independently conduct such analyses using free software and connect the computational results to archaeological/scientific interpretations related to the phenomenon in question.

Target Audience

All interested doctoral candidates at BerGSAS, including those from disciplines other than archaeology. Active participation will be certified. Depending on the study regulations, it may count towards the curriculum of the BerGSAS doctoral programmes (please check this with the programme coordinator beforehand).

Prerequisites

No specific prior knowledge is required, though an interest in and willingness to learn abstract methods are essential. Participants should bring their own laptop with the free software—available at https://www.nhm.uio.no/english/research/resources/past/ (PAST = Palaeontological Statistical Package)—already installed. Alternatively, participants may use the PCs available in the course room.

Registration and inquiries

Registration & inquiries regarding general organization: regina.attula@berliner-antike-kolleg.org
Inquiries regarding content: georg.roth@fu-berlin.de

Introductory literature

Thomas Wickens, The Geometry of multivariate statistics (Hillsdale 2014) <Primo: SK 830 W636>
Michael Greenacre, Correspondence Analysis in Practice (Boca Raton 2017 2.) <Primo: SK 840 G798(3)>
Pierre Legendre and Louis Legendre, Numerical Ecology (Amsterdam 2012) <Primo: EBOOK>