Rose diagrams and statistical tests for archaeological orientation data

A case applied to the Southfleet Road Elephant site, Ebbsfleet, UK

Bibliographic Data

ID11778670
AuthorsRichard J Hewitt (0000-0003-4169-8647), Simon Parfitt (0000-0003-2660-9570, Natural History Museum), Simon A Parfitt, Francis F Wenban-Smith, Francis Wenban‐Smith (University of Southampton)
Year2018
Publication date2018-12-14
Open AccessNo
TypePREPRINT
PublisherCenter for Open Science (PUBLISHER)
DOI10.31235/osf.io/f2vzs
OpenAlexW4211099732
LanguageEN
Citations received2
References cited7

Archaeologists and other fieldworkers frequently record the orientation of artefacts or faunal remains in order to understand depositional and post-depositional processes. However, analysis of this kind of data is complex and not always well-understood by investigators. For example, rose diagrams are commonly used to display orientation data. While rose diagrams are a useful and intuitive way of visualizing the data, they are not on their own sufficient except in cases where preferential orientation is very evident. Class frequency, rather than the square root of class frequency, is commonly used to interpret the results. This can give a misleading impression of preferential orientation, when in fact observed patterns do not deviate significantly from random. Conversely, data may deviate significantly from random but this may not be clear from the diagram. At the same time, classic tests like the chi-square test are not appropriate for the display of circular data and alternative approaches are needed. This paper discusses the history of the use of rose diagrams to display orientation data, briefly reviews the archaeological literature in this regard, and discusses key problems and limitations. We then make some recommendations for best practice, which we demonstrate using R software for the Lower Palaeolithic “elephant butchery site” at Southfleet Road, Ebbsfleet, UK, excavated in 2004. The methods employed are clearly explained and all data are provided to ensure that the procedure is easily repeatable without specialist knowledge or programming experience

Archaeology · Class (philosophy · Geography · Geometry · Orientation (vector space · Rose (mathematics · Computer Science · Image Processing and 3D Reconstruction · Mathematics · Morphological variations and asymmetry · Pleistocene-Era Hominins and Archaeology · Artificial Intelligence · Geology

Tools

Open DOIOpen Access