Geographical distribution
The geographical distribution of finds is not even. The uneven distribution may be real and reflect the numbers of rings that were made and/or lost in those particular regions but it may be due to the non-random nature of the records on the PAS database. A detailed study of items recorded on the PAS database by Katherine Robbins (2012) showed that bias may be introduced at any of seven stages from item deposition through to its subsequent recovery; these stages are burial/loss, preservation, survival, exposure, recovery, reporting, and recording. Some of these factors such as preservation, survival and exposure are unlikely to bias the records for gold rings but the remaining factors may influence the distribution of finds.
Taking these factors in turn, rings interred during a burial are very unlikely to be discovered whilst metal detecting as they would be beyond the detection reach of standard metal detectors and since the overwhelming majority of rings recorded on the PAS database were found using a metal detector this factor can be excluded as a source of bias. Casual loss of a ring may occur at any time and in any location although it might be assumed that locations used by a larger number of people e.g. inhabited areas, public footpaths etc might result in a larger concentration of finds than random open areas such as farmland. It might therefore be assumed that inhabited areas are more likely to be targeted by metal detectorists; however, previous and current ownership of the land and willingness of the landowner(s) to allow metal detecting will influence whether lost rings in such areas might be found. Furthermore, many metal detecting club ‘digs’ and rallies are held on land that is chosen on the basis of sufficient area to accommodate large numbers of individuals rather than the presence of footpaths etc and such digs/rallies with large numbers of detectorists therefore potentially searching land more efficiently than sole detectorists might be able to do thus making the discovery of rings more likely. In one of the authors (DLA) limited experience only one of the four posy rings found by them was on or alongside a public footpath with the remaining three rings being found on farmland with no existing public footpath and no history of a footpath on maps dating back to 1885. The geographical location of potential finds therefore is quite a complex factor to assess in terms of recording bias.
Recovery of artefacts is also a complex area to assess and is dependant on a number of factors. The type of metal detector used will significantly influence the detection of artefacts, with some machines being more sensitive than others, particularly to gold; they are therefore able to detect at greater depths. Some have better discrimination abilities are able to differentiate different types of metals more readily. The expertise of the operator in using the machine correctly and to the maximum of its abilities will affect the interpretation of the signals being received and can nullify any advantages of particular machines. Ground conditions (presence of minerals, moisture content etc) will also potentially influence the ability to detect buried objects. It can be seen that, again, this is a complex area of potential bias.
Whilst there is a legal requirement to report items that are suspected to fall within the remit of the Treasure Act unfortunately not all individuals are aware of this and do not report or record finds and, sadly in some cases, some unscrupulous individuals simply ignore the requirement. It is not known how many items do not get recorded nor whether such items tend to be the more valuable items.
Lastly, the presence or absence of a Finds Liaison Officer in particular regions, the date at which that region started to submit data to the PAS database will also potentially affect numbers of artefacts recorded.
Nonetheless, despite there being a wide range of factors that potentially might influence recording bias one might assume that such biases would affect all types of inscribed gold rings equally. However it can be seen from the figure below that there do seem to be geographical differences (more on this when the analysis is complete). The Eastern counties of England do seem to be particularly productive in terms of numbers of rings reported.