less detail |
## 4.5 Percentage of artefacts per siteEven without executing a multivariate statistical analysis, a number of problems can be pointed out here. Quite remarkable in Table 8, section 4.1 are the small numbers of artefacts: a specifically archaeological problem! Both the total of artefacts and the number of guide artefacts are extremely low. There are a relatively large number of sites containing only a single guide artefact, without containing many finds from other categories of the same period. There are only four sites of significant size, one of which towers far above the others, with several hundreds of artefacts. When a multivariate classification is executed more or less automatically, the small numbers have a large impact. A site assemblage which consists 100% of pointed blades, must be the result of some very strange activity! This may be an extreme example, but due to the distorting factors (3.1) described above, the absolute numbers of artefacts per site are not stable. Site 52B-168, not one of the smallest clean Michelsberg sites, may serve as example here. This site assemblage consists of 1 macrolithic artefact, 2 scrapers and 4 other artefacts (mainly production debris). This can be expressed in percentages as: 14% macrolithic artefacts, 29% scrapers and 57% other artefacts. A single new macrolithic artefact significantly changes those percentages to 25%, 25% and 50%, respectively. Due to the small size of the sites, both the absolute numbers and the percentages are not very stable and therefore archaeologically less suited for statistical analysis. After all, a new visit by an archaeologist to the site may drastically change the numbers. |

© Internet Archaeology
http://intarch.ac.uk/journal/issue4/wansleeben/noprcnt.html

Last updated: Wed Feb 25 1998