Glossary

Cluster or nested sampling Divides the project area into arbitrary primary clusters that can be subdivided again for the selection of sampling units.

Cultural Resource Management (CRM) CRM is the

American practice of performing archaeological investigations prior to ground-disturbing construction so that archaeological resources, values, and information are preserved for future generations. It is sometimes referred to as public archaeology.

In America, this practice is driven by the amended National Historic Preservation Act of 1966 and by the National Environmental Policy Act of 1969. ditch The low area surrounding an earthwork fortification and the dirt from its construction excavation is ‘thrown up’ to form the interior parapets. The geometry of the ditch and the parallel parapets defines the fortification, for example, a hexagonal star, rectangular, etc.

Glacis The outer plain that surrounds, and extends beyond, the ditch and which is cleared of vegetation so that an attack can be detected early.

Parade The central space of an earthworks fortification and is protected by the elevated parapets. parapets The elevated perimeter of an earthworks fortification from which quartered militia can fire at approaching attackers. Parapets define, protect, and separate the inner, parade portion of fortifications from the exposed glacis on the exterior side. primary cluster Refers to cluster sampling and to the initial creation and location of a block of space, the primary cluster that should be a microcosm of the variability in the target or sampled population. In nested or multistage sampling the primary cluster is subdivided several times into sub - or sub-sub-clusters before the desired sampling units are isolated and selected. project area A nonsampling, colloquial term for a site or region in which research, usually fieldwork, will be conducted, corresponding under certain sampling conditions to the target population.

Planning Policy Guidance, no. 16 (PPG 16) Guidance issued in November, 1990, by England’s Secretary of State for the protection of archaeological remains and monuments when they are threatened by development and construction. PPG 16 outlines the necessary procedures and organizations involved in archaeological protection. It appears to be a more detailed English equivalent of the American Section 106 process under the National Historic Preservation Act (see CRM in Glossary). research design A written plan of the archaeological research to be accomplished and the methods to be employed.

Sample size/fraction The number of sampling units selected for data collection and usually expressed as a fraction or percentage of sampled population. sampled population The larger area from which smaller sampling units are selected for data collection and analysis. It is nearly synonymous with the frame as a list of all sampling units that are accessible to the researcher for investigation and study. sampling element The smallest unit of observation and measurement for a given research design and objective.

It is substantive and cultural in nature, as distinct from the completely methodological sampling unit. sampling unit A spatial unit of survey or excavation that is defined by its shape (quadrat or transect, rectangular, linear or square) and by its size (5x5ft or 100 sq. ft, or a 10x500 ft transect). The size of the sampling unit is not to be confused with sample size/fraction (see above). simple random sampling (SRS) The most basic kind of sampling that insures that each equal-sized, non-overlapping sampling unit has an equal probability of selection into the overall sample for data collection and eventual analysis. stratified sampling The least arbitrary sampling scheme in which the project area is divided into culturally or environmentally meaningful zones that are called strata from which sampling units are chosen (by SRS or unaligned) for investigation.

Subclusters and sub-sub-clusters Used, respectively, in the two - and three-stage cluster sampling schemes when the primary cluster is divided twice and three times, respectively, before reaching the level of sampling units. In two-stage cluster sampling, the sampling units are selected randomly (SRS) from the subcluster. In three-stage sampling, the sampling units are selected randomly (SRS) from the sub-sub-cluster. The unaligned technique may be used in lieu of SRS, but the systematic variant should be avoided.

Systematic sampling A variant of SRS that results in sampling units being located at randomly selected, equally spaced intervals across the sampled population.

Target population The largest spatial area or conceptual topic for which the interpretations of research are statistically valid.

It is the sampling equivalent of the project area under certain conditions described in the text.

Sampling is a methodological tool of selection, discovery, and analyses used consciously and unconsciously by archaeologists in library, fieldwork, and laboratory work at all levels. Formal, quantitative sampling is based on probabilistic theory which is also associated with statistics. Therefore, this essay will include a limited statistical discussion, but focus mainly on the concepts and the mechanics of sampling, using actual survey and excavation projects, supplemented with experimental sampling, as illustrative examples. In preparation for fieldwork, certain publications selected for study help to create the research design, while other publications are not studied. During the discovery phase of fieldwork, space is sampled during survey and excavation to discover unknown cultural and environmental data about a region, site, feature, or artifact scatter, while some other data-bearing spatial areas are not selected for field investigation. During laboratory work, certain features and artifacts are selected for further study, while certain others are not selected for analysis. Of the selected features, artifacts, and ecofacts, certain attributes/variables are further selected for study from many possible attributes, while other attributes/variables are not selected, depending on the project’s research design. During report writing, certain attributes of the many recorded data classes are selected for textual description and interpretation, while others never make it into the final report. At all levels, archaeological research is a continual sifting and selection process, a winnowing down, many times over, to separate the proverbial chaff from the grain, a winnowing down from concepts and research objectives to objects and three-dimnensional space to qualities of the newly discovered data and to the past cultural meanings of all of the above. At any one, or all, of the above levels in the hierarchy of archaeological research, this selection process has been referred to as sampling. Sampling is beautifully suited to archaeological research because it is also hierarchical, as will be discussed herein. At the outset, it should be stated that complete coverage in library/archival research, survey, excavation, laboratory analysis, and report writing is the ideal and is usually preferable to sampling. However, the real-world constraints of costly and time-consuming, labor-intensive archaeology frequently demand the use of sampling as part of the archaeologist’s tool-kit.