Sampling in eHRAF

It is rarely necessary to use the entire eHRAF World Cultures database for comparative studies. As long as a sample is representative of some well-defined universe of cultures, societies, or archaeological traditions, sampling a sub-set saves time and effort.

Researchers wishing to have a sample that claims to be representative of the cultures of the world should consider some of the eHRAF sub-samples, such as the 60-culture Probability Sample Files (PSF). A simple random sample can be used to supplement the PSF. In addition, eHRAF has all 186 societies in the Standard Cross-Cultural Sample (SCSS).

eHRAF World Cultures contains four samples that can be considered representative:

Ethnographic Atlas Sample (EA): more than 1,264 societies pinpointed in time and (usually) space
Probability Sample Files (PSF): 60 largely preindustrial societies that meet certain data quality controls, one randomly chosen from each culture area.
Simple Random Sample (SRS): currently 28 societies randomly chosen from a compiled list of over 8 cross-cultural samples.
Standard Cross-Cultural Sample (SCCS): consists of 186 anthropologically described societies pinpointed in time and space.

For eHRAF Archaeology, HRAF has employed two types of sampling. The first is simple random sampling (SRS) from the Outline of Archaeological Traditions. The second is to include entire tradition sequences, usually three or more archaeological traditions.

Culture Level Samples and Document Level Samples

When performing a search or viewing the Browse Documents page, both Culture Level Samples and Document Level Samples are available in filters. This is what they mean:

Culture Level Samples: includes every document related to every culture in the chosen sample
Document Level Samples: these documents are tagged by HRAF staff as relating to cultures in the relevant sample

Note that eHRAF Archaeology does not have Document Level Samples.

Regarding document level samples, to facilitate research for those wanting to code their own variables using eHRAF World Cultures for the EA and SCCS samples, the HRAF staff have flagged those documents that, in HRAF’s judgment, provide an appropriate match to the time and place foci of the EA and SCCS. Note that in almost all cases, eHRAF World Cultures contains many more documents than those flagged. This is because the HRAF collections try to cover multiple times and places to facilitate understanding of change over time and of regional variation.

Entire, Partial, and Cautionary

Document Level Samples use the terms Entire, Partial, and Cautionary. This is what they mean:

Entire: an entire match, where these documents are about the culture in the time and place specified by the EA or SCCS
Partial: a partial match, usually because the document covers multiple time periods or multiple cultures
Cautionary: the time or place of the document is somewhat ambiguous and we urge caution when using the source.

If a culture is a part of the EA or SCCS but a document about that culture is not part of a document level sample, it may mean that the document is not a time and place match. We still include these results in order for researchers to have access to additional data that may be relevant to them despite not being an exact EA or SCCS match.

How to find and narrow by sample type(s) in eHRAF

You can filter by sample type in eHRAF while on the Browse Document or Browse Culture pages or while performing a search. On any of those pages, click the Show Filters button to open the filters sidebar on the left and then select the sample filters you wish to apply.

The Basic Guide to Cross-Cultural Research on the HRAF homepage contains further discussion about sampling including suggested reading.