I think it is necessary to keep in mind that accuracy is relative and time dependent. Until the advent of State Plane Coordinate system mapping tied to CAD systems, accuracy was eyeballed most of the time using the standard USGS Quads. Absent landmarks or natural features, ±200 feet was normal. No one seems to question the accuracy of the base map, but once one tries to edge map USGS Quads, 300 foot differences are not unusual. In the real world, you could theoretically stand with one foot 300 feet from the other. Tall folks may have longer strides, but not that much so;).

SPC systems with 1-5 foot contour mapping allow for greater accuracy, for sure. But even that is a judgment call and will have at least, for eyeballed locations, a ±20 foot fudge factor.

GPS readings have an EPE that is based on worst-case scenarios. I have used hand-held Garmin GPS and gotten identical readings from the same benchmarks on 3 consecutive days for two of the three with the fourth being a USGS benchmark that was hidden amongst some cedar trees.

The other issue is whether the testing regimen is accurate enough to provide a true edge. Shovel testing is minimally accurate as compared with visual examination of the surface of a field with no vegetation. Slightly less accurate, but still more-so than shovel testing, are prepared strips at the same interval as shovel tests that allow for an interpolated site edge. The problem has been and will continue to be determining the actual edge of the site with whatever method one chooses.

Ambiguity, once accepted, is part and parcel of measurement of anything. And, I suppose, the final issue is does absolute accuracy have any real-world meaning. We can measure in 100ths of a foot using EDM's but whether what one is measuring to that accuracy is the edge is the ambiguity. The least accurate part of the system governs the overall accuracy of the system in which it is used.

One can waste enormous amounts of time and money attempting to achieve accuracy that is in the end meaningless without understanding the system on which it is based, and without understanding the system in which it is used and their fundamental ambiguities.

Lyle Browning, RPA


On Nov 11, 2012, at 3:53 PM, Jim wrote:

> Fellow HistArchers:
> I am a member of a committee established by the Council for Maryland Archeology charged with the development of a white paper on best practices. My particular charge is to identify the best practices in accurately and precisely documenting the locations and boundaries of archaeological sites.
> I have examined the standards and guidelines published for eight eastern states and one western state in the USA. The words 'accuracy' and 'precision' are mentioned in only one or two and none offers guidance on the matter other than investigators documenting the instrument and standard error when using GPS units. I'm seeking input on best practices in mapping site locations, boundaries, units, etc., with or without GPS units. Discussion welcome.
> Jim Gibb
> 
> James G. Gibb
> 
> Gibb Archaeological Consulting
> 
> 2554 Carrollton Road
> 
> Annapolis, Maryland USA ?? 21403
> 
> 443.482.9593 (Land) 410.693.3847 (Cell)
> 
> www.gibbarchaeology.net ? www.porttobacco.blogspot.com